(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
See other formats

Full text of "Plato's use of fallacy,: a study of the Euthydemus and some other dialogues"




Digitized by the Internet Archive 

in 2012 with funding from 

LYRASIS Members and Sloan Foundation 



A Study of 

the Euthydemus and Some 

Other Dialogues 




First Published 

in the United States of America 


Copyright Rosamond Kent Sprague 1962 


Printed in Great Britain 

A. L. P. 

€l86tl <f)Ojri 


my thanks are due to the American Association of University 
Women who awarded me the Clara A. Scott Fellowship for the 
year 1 956-1 957. As a result I was able to work on the subject 
of this book at Cambridge University with Dr. A. L. Peck of 
Christ's College, and at Princeton University with Professor 
Gregory Vlastos. I am deeply grateful to both of these scholars 
for their generous gifts of time and sympathy over a period of 
several years. I wish also to thank Dr. J. M. E. Moravcsik of the 
University of Michigan, Dr. Hugues Leblanc of Bryn Mawr 
College, and Dr. Martin Ostwald of Swarthmore College, all 
of whom have been kind enough to read the appendix, and a 
number of other scholars who have taken time to talk with me 
about various aspects of my work, especially Professor Harold 
Cherniss of the Institute for Advanced Study, Professor W. K. 
C. Guthrie, Mr. J. R. Bambrough and Mr. R. L. Howland of 
Cambridge University, Dr. R. S. Bluck of the University of 
Manchester, and Miss A. N. M. Rich of Swarthmore College. 
The English translations of the various passages from Plato 
and Aristotle are from the Loeb Classical Library, as are the 
quotations from the Greek, by kind permission of the Harvard 
University Press and William Heinemann. 






Introduction 271A-275D 1 

Scene I 275D-278E 3 

&«!*// 278E-282D 8 

Scene III 282D-288D 12 

Scene IV 2WD-2MK 20 

Scene V 293A-304C 22 

Epilogue 304C-307C 31 


III CRATTLUS 429Bff. 46 




there is no doubt that there are many fallacious arguments in 
Plato's dialogues. This book is an attempt to try out the hypo- 
thesis that Plato was fully conscious of the fallacious character 
of at least an important number of these arguments, and that 
he sometimes made deliberate use of fallacy as an indirect 
means of setting forth certain of his fundamental philosophical 

Of course there may well be arguments in the dialogues which 
Plato regarded as sound but which are in fact fallacious. It is 
not my purpose to deny this nor to examine passages where 
Plato may have committed unintentional logical errors. My 
purpose is rather to insist that in the case of certain specific 
fallacious arguments Plato was fully aware of the fallacy and 
used it for a purpose. 

As for what is meant by speaking of the way in which Plato 
'uses' certain fallacies, the view that I shall develop may be 
summarized as follows: Plato's general method in these cases 
is, not to incorporate or build fallacious arguments into his 
dialogues in such a manner as to produce conclusions based 
upon unsound foundations, but rather to introduce them for the 
purpose of working out their implications. Thus he is able (i) to 
expose them for what they are, and sometimes (2) to clear 
away possible lines of attack upon his own position, or even (3) 
to show that when the proper correction is applied his own 
views receive support. 

The book, then, has been called Plato's Use of Fallacy because 
it is this latter point, the connection between fallacy and 
Plato's philosophical position, which has engaged my attention 



most. But although this subject comes first in importance it 
cannot come first in order since it is clearly not reasonable to 
begin to discuss how Plato has used a particular fallacy unless 
we have previously identified that fallacy. Further, if it is to be 
maintained that such use was deliberate, the case for conscious- 
ness must have been settled in advance. 

From the point of view of method, therefore, it seemed to 
me essential to begin with a dialogue in which a fair number of 
fallacies could be readily identified, and the controversial 
problem of consciousness could be quickly settled. The dialogue 
which satisfies these conditions best is clearly the Euthydemus. 
The two sophists who perpetrate practically all of the numerous 
fallacies contained in the dialogue are obviously the villains of 
the piece, whereas Socrates and Gtesippus are equally obviously 
the heroes. It has not, to my knowledge, ever been suggested 
that Plato looked on any of these arguments as valid; the 
dialogue has always been regarded as a piece of scathing satire 
designed to expose a variety of eristic tricks. What is really 
surprising is that when similar arguments occur in other 
dialogues their likeness to arguments in the Euthydemus passes 
unnoticed, or, if it is noticed, the possibility that Plato has 
made deliberate use of such arguments is not considered 

I therefore begin with a summary and analysis of the entire 
Euthydemus. (I have preferred to deal with the dialogue as a 
whole rather than to excerpt the passages in which fallacies 
specifically occur, first, because the dialogue is relatively un- 
familiar to students of Plato, and, second, because I believe 
that these passages are best understood if taken in their proper 
setting.) Using the Euthydemus, then, as a species of control, I 
have next gone on to discuss important passages in two addi- 
tional dialogues, the Theaetetus and the Cratylus. The choice of 
these passages was dictated by the fact that I wished to con- 
centrate on philosophical points and thus to avoid taking time 
to argue the case for consciousness afresh in each instance; 
therefore I decided to choose fallacies whose similarity to 
fallacies in the Euthydemus was close enough to make this 
unnecessary. In other words, I take it as legitimate to assume 
that if Plato understands e.g. the fallacious argument against 
false speaking in the Euthydemus, he understands it in the 



Cratylus also. The final chapter, on the Hippias Minor, is a 
departure from the general scheme. In this dialogue there 
occur fallacies which are not precisely similar to arguments in 
the Euthydemus, but these may, I believe, be regarded as 
conscious from independent internal evidence. 

In all three of these chapters the identification of the various 
fallacies discussed and the decision that their use is conscious is 
only a preliminary procedure; the major part of each is devoted 
to an attempt to come to some conclusions as to what may have 
been Plato's purpose in the employment of these arguments. 
In each case it will appear that Plato, when he makes use of a 
fallacious argument, does this in order to call our attention to 
some fundamental philosophical problem. In this respect he 
differs radically from the eristic sophist who uses fallacy for the 
single purpose of achieving victory in the war of words. 

The choice of the Euthydemus as starting-point has very 
naturally affected the variety and type of fallacy considered. 
The sophists in that dialogue are neo-Eleatics, that is, their 
arguments (which can in the main be reduced to two, equivo- 
cation and secundum quid) are based upon the philosophical 
position of Parmenides. As a result the clash between Eleatic 
monism and Plato's own view is a recurring theme in the first 
three chapters. The Hippias Minor, however, is concerned not 
with the difficulties in Eleaticism but with some ambiguities in 
the Socratic ethics ; thus a chapter on this dialogue has extended 
the range of problems which I believe that Plato was concerned 
to elucidate by deliberate use of fallacy. 

The number of dialogues studied is a small one, but as this is 
very much an experimental essay, it seemed best to keep to 
passages for which it was possible to make a really strong case 
for conscious use. (Passages from additional dialogues have, 
however, been discussed briefly in the conclusion, the appendix, 
and the notes.) On the other hand, the number of philosophical 
problems which arise is by no means small: it includes, for 
instance, the status of particulars, the nature of participation, 
the distinction between otherness and not-being, the claims of 
Parmenidean monism and Protagorean relativism to account 
for learning and memory, the difficulties in resemblance as the 
basis for a theory of language, and the meaning of ethical 
terms like 'good' and 'voluntary'. In other words, the study of 



even a small group of fallacious arguments in Plato can be seen 
to lead straight to the heart of his philosophy. (Such a study 
may lead also to interpretations of particular dialogues which 
are not completely orthodox: for instance, I have become 
convinced that the Cratylus has much more significance for the 
theory of Forms than is oftentimes supposed.) 

I have also added an appendix dealing, in an adversely 
critical way, with some of Fr. I. M. Bochenski' s remarks on 
Plato's logic. I have done this, not from a spirit of controversy, 
but because Fr. Bochenski seems to me, in work which is bound 
to have its effect upon historians of logic, to have underestimated 
Plato's logical competence very gravely. As far as one can judge, 
his procedure seems to have been to collect from the dialogues 
a number of passages containing logical errors and to offer 
these as evidence on the basis of which we are invited to form 
an opinion of the extent of Plato's contribution to logic. But 
although it is useful to have our attention called to passages in 
which fallacies occur, it is extremely misleading to suggest that 
an opinion of Plato's logical skill should be based on such 
passages in isolation: to do so is to place the dialogue, as a 
literary form, in the same class as the prose treatise. In contrast 
to Bochenski, therefore, I have usually tried to study Plato's 
fallacious arguments in the context of whole dialogues, and to 
give at least some attention to the character of the particular 
person with whom Socrates happens to be conversing when the 
fallacy occurs. As the reader will see, the result has been that 
my estimate of Plato's logical powers is unusually high. 

It should also be mentioned that although the Sophist is a fine 
hunting-ground for arguments which also appear in the 
Euthydemus, and although illustrations from this dialogue would 
have increased the temporal range of my examples, I have not 
attempted to use the Sophist in view of the fact that a number of 
these parallels have already been noted by Dr. A. L. Peck in 
his paper 'Plato and the METIETA rENH of the Sophist: 
a Reinterpretation' {Classical Quarterly, N.S., vol. II, Nos. i and 
2, January-April, 1952, pp. 46-7). It was Dr. Peck's paper 
which first led me to study the Euthydemus in detail, and I feel 
indebted to this article for the notion that some, at least, of the 
fallacies in Plato's dialogues might be part of a deliberate plan 
and not the errors of a logical incompetent. 



I hope, then, that this book will call attention to the Euthy- 
demus as an important but neglected dialogue, and that it will 
serve as a defense of Plato's understanding of a number of 
logical points ; above all I hope that it will help to recommend 
the study of fallacy in Plato as a possible approach to the inter- 
pretation of his dialogues. 



The Euthydemus 

from the point of view of pure symmetry, the Euthydemus is 
probably Plato's most beautifully constructed dialogue. A brief 
introductory conversation between Socrates and Grito sets the 
stage as the dialogue begins. Socrates then relates to Crito a series 
of five alternating scenes, of which the first, third, and fifth are 
exhibitions by two sophists, Euthydemus and Dionysodorus, and 
the second and fourth are exhortations to philosophy by Socrates 
himself. (The fourth contains an interruption by Crito, which 
serves to vary the routine.) A final exchange between Socrates 
and Crito concludes the dialogue. 

Introduction, 271A-275D 

In the early part of the dialogue we have a number of hints 
both as to the intrinsic importance of the matter in hand, and 
to the probable inability of Euthydemus and Dionysodorus to 
deal with it adequately. 

Socrates had been on the point of leaving the palaestra on the 
previous day when he was prevented by his daimonion 272E.* 
Socrates' spiritual sign does not usually occur in circumstances 
which are merely trivial, so that it seems unlikely that Plato 
would represent the daimonion as intervening to make the 
discussion possible unless what followed were to be of real 
consequence. The reason why Socrates has been detained is 
speedily made clear by the claim of the sophists to transmit 
virtue (dperrjv 7rapa8ovvai) at 273D, since this claim immedi- 

* In the case of a single Platonic reference preceded by no mark of 
punctuation, I have regularly omitted parentheses. 
B I 


ately raises two of the most crucial Socratic questions: can 
virtue in fact be taught, and, if so, who are the teachers of it ? 
If eristic had been simply a game, it would not have been 
worthy of Plato's serious attention, but since its practitioners 
posed as teachers of virtue, it was a contender for the same 
activity as his own dialectic, and needed to be exposed in all its 
shallowness and frivolity. The contrast between the two is 
subtly implied by Socrates at 274D: 1 he asks Dionysodorus 
whether the art that sets up to teach virtue (in this case eristic) 
can also persuade the pupil that virtue is teachable and the 
persons teaching it (in this case the sophists) are the ones from 
whom he should learn, or whether some other art be required 
for this purpose. Dionysodorus replies confidently that no other 
art is needed, but we can guess that if any art is capable of 
settling these matters, it will be dialectic, not eristic. This is 
exactly one of the differences between them, that dialectic 
contains its own justification, 2 whereas eristic does not and 
cannot. The sophists will never, by means of their own art, be 
able to show either that virtue is teachable or that they are its 

1 Strictly speaking, this sentence should read ' Plato represents Socrates as 
subtly implying the contrast between the two at 274D' and so forth. In 
order to avoid circumlocutions of this sort, which would be cumbersome 
and repetitious, I have spoken throughout of Socrates only. This is, of course, 
somewhat misleading, but will not, I hope, be misunderstood. It has not 
been my purpose in this book to distinguish the historical from the Platonic 
Socrates. My assumption is that whatever happens in the dialogues (with 
respect to Socrates or any other speaker) was at least understood by Plato. 
Whether or not he originated or even believed the particular point at issue is 
quite another matter, and one with which I shall not often be concerned. 
If, then, I speak of Socrates as, e.g., appearing to anticipate that Euthydemus 
will say so and so, I am not meaning to assert that the historical Socrates 
need ever have done such a thing or need even have been likely to do it. I 
am simply employing a short-hand method of referring to something which 
I feel that Plato depicts Socrates as doing at a particular stage in a particular 

2 By saying that dialectic contains its own justification, I mean that 
Plato appears to regard this art as able, without the aid of any auxiliary art, 
to bring a person to the point at which he conceives it necessary to choose 
and follow virtue. We see this in the case of Cleinias, who, when he has for a 
short time acted as answerer in a conversation with Socrates, is convinced 
not only that virtue (here wisdom, but see below, p. 12) is teachable 282C, 
but that it is necessary to pursue it in order to be happy 282D. When, on 
the other hand, Cleinias acts as answerer for the sophists, he reaches no such 



teachers ; thus their so-called teaching has no basis on which to 

The contrast between dialectic and eristic is, in fact, a 
recurring theme throughout the entire dialogue. The most 
obvious and striking form in which this contrast occurs is in the 
dramatic device of the alternating scenes, but it is evident also in 
numerous small details. Socrates' interest in the future of 
Gleinias 275B is a genuine one, for instance, whereas the sophists 
receive the candidate indifferently; they do not care what 
happens to Cleinias so long as he is willing to act as answerer for 
the demonstration. In other words, dialectic (in the person of 
Socrates) is represented as an art whose practice displays a real 
concern for the welfare of the individual soul, whereas eristic 
(in the persons of Euthydemus and Dionysodorus) appears to 
be completely lacking in any such concern. So long as dialectic 
and eristic are in conflict over this or comparable points, 
eristic cannot be dismissed as trivial or as simply a joke. It is in a 
sense a joke, but a joke with dangerous pretensions. 

The first hint that Euthydemus and Dionysodorus will not be 
able to come to grips with whatever problem may be proposed 
occurs earlier, in the course of the introductory conversation 
with Crito. Here Socrates describes the two brothers as having 
acquired the skill of confuting any argument 'equally well 
whether it be true or false ' (ofiotcos idv re i/jevSos idv re dXrjdes fj 
272A). In other words, they are not really interested in truth 
but only in verbal battles (iv rots Xoyots fxax^aOai 272A). 
Socrates' low opinion of the methods of the sophists is also plain 
from his comments on the fact that they have been able to 
learn their skill so quickly and at such an advanced age 272C. 
(The facility with which this kind of questioning is acquired is 
noted at a number of points in the dialogue, e.g., 300D, 303E, 
304A.) Thus Plato's attitude towards the sophists and their 
facile technique is clear even before he mentions their claim to 
be teachers of virtue at 273D, and we are well prepared for the 
failure of this technique to achieve its object before the demon- 
stration begins at 275D. 

Scene I 275D-278E 

Euthydemus' first question to Cleinias, whether the learners 
are the wise or the foolish, comes as a slight anti-climax 



(no doubt designedly so) after Socrates' invocation to the Muses 
and Memory for assistance in recalling such enormous know- 
ledge 275D. Socrates encourages Cleinias to answer bravely, 
since perhaps the sophist is doing him the greatest possible ser- 
vice 275E. Dionysodorus, however, whispers to Socrates that 
the boy will be refuted no matter how he answers. 3 (Here again, 
Socrates is genuinely concerned with whether Cleinias will 
benefit from the demonstration, while the sophists are using him 
simply as a means of displaying their own dexterity in argument.) 
In answer to the first question, Cleinias replies that it is the 
wise who are the learners 275E. Euthydemus, however, points 
out that when Cleinias was learning from his teachers he did not 
yet know the things he was learning. Therefore, since he did not 
know these things, he was foolish, not wise, and his present 
answer was wrong. Dionysodorus immediately goes on to ask, 
which boys learnt the piece dictated by the writing-master, the 
wise or the foolish? (276C) Cleinias answers, the wise, and so 
contradicts his previous statement. Euthydemus continues with 
yet another question, which, as Dionysodorus confides to 
Socrates, is like the others in admitting of no escape : do the 
learners learn what they know or what they do not know? 
(276D) Cleinias answers that they learn what they do not know. 
But does Cleinias know his letters ? Yes, of course. Then when 
anyone dictates, he dictates what Cleinias knows already, since 
dictation consists of letters. If Cleinias learns anything from 
dictation, he learns what he knows, not what he does not know, 
so he is wrong again. But, says Dionysodorus, to learn is to 

3 The reason why Cleinias will be refuted (i.e., sophistically refuted) no 
f matter which answer he gives, is that a question which contains an equivocal 
term (as we shall see that this question does), is really a form of the fallacia 
plurium interrogationum (e.g., have you stopped beating your wife?). Aristotle 
is instructive on this point : ' If one does not make two questions into one, 
the fallacy which depends on equivocation and ambiguity would not 
exist either, but either refutation or absence of refutation. For what is 
the difference between asking whether Callias and Themistocles are musical 
and asking the same question about two people both with the same name ? 
For if one indicates more things than one, one has asked more questions 
than one. If therefore it is not correct to demand simply to receive one 
answer to two questions, clearly it is not proper to give a simple answer to any 
equivocal question, even though the term is true of all the subjects, as some 
people claim one ought.' {On Sophistical Refutations, 175b 39 ff., trans. 

\ 4 


receive the knowledge of that which one learns, and to know is to 
have knowledge already. Gleinias agrees to this and to the next 
assertion, that not knowing is not yet having knowledge. 
And are those who receive anything those who do or do not 
have it already? Those who do not have it, answers Cleinias, 
and is refuted again, since Dionysodorus replies that the 
learners are those who receive knowledge, not those who have 
it, and are thus the same as those who do not know. 

Euthydemus is about to embark on another round of ques- 
tions when Socrates intervenes 277D. He explains to Cleinias 
that the sophists are simply preparing him for his initiation 
by these merry-makings. They want him to find out about 
the correct use of words (ovofxdrwv opdorjjs 277E) and are 
pointing out that 'learning' (to fiavdaveiv) is used both of 
someone who, having no knowledge of a thing at first, acquires 
it later, and also of someone who does have knowledge of a 
thing and uses this knowledge to go on investigating the same 
thing further. In other words, they want Cleinias to notice that 
the same word is used for people who are in the opposite 
states of knowing and not knowing. Socrates calls such questions 
frivolity (iraLhid 278B), since even if one were to learn many 
or all such tricks one would know nothing more about how 
matters really stand (rd fiev Trpdy^iara oz>8ev av [xdXXov eiSeirj 
77-77 exec 278B). The gentlemen will probably soon give up this 
sport and get down to the serious business of exhorting Cleinias 
to the pursuit of wisdom and virtue. First, however, Socrates 
will give them an example of how he himself would like to see 
the discussion conducted. 

This speech of Socrates (277D-278D) is notable on two 
counts: it explains the fallacious argument which the sophists 
have used to refute Cleinias, and it gives us Plato's funda- 
mental objection to such arguments. 

All of the refutations in this first scene (refutations which are, 
of course, not real but apparent) turn upon the fact that a 
man who learns is in one sense ignorant (that is, he is ignorant 
of the things he is about to learn) and in another sense wise 
(that is, he has sufficient intelligence to be able to add the new 
piece of information to his previous stock of knowledge) . Thus 
Socrates is pointing out that 'learn' is an equivocal term and 
that the sophists are here employing the fallacy of equivocation. 



The contradictions result from the fact that this equivocal 
term can be used with respect to persons who are in opposite 
conditions, who are either knowing or not knowing, as 
Socrates says at 278A. The sophists are clearly using not only 
the fallacy of equivocation but also another fallacy closely 
connected with it, the one known traditionally as a dicto 
secundum quid ad dictum simpliciter* This fallacy consists in taking 
absolutely what should be taken only accidentally, e.g., to go 
from 'knowing one's letters' to simply 'knowing' (which can 
also be taken as equivalent to 'wise') or from 'not knowing the 
things one learns ' to simply ' not knowing ' (which can also be 
taken as equivalent to 'ignorant' or 'foolish'). The two 
fallacies in combination give the sophists a very effective 
instrument for refutation of the unwary. In the first place they 
take pains to phrase their initial questions so that Cleinias will 
have to choose between a pair of exhaustive alternatives. The 
equivocal nature of ' learn ' then enables them to produce the 
opposite of whichever one he chooses. If he says the learners are 
the wise, or those who know, they bring forward a case in which 
those who learn are ignorant of some one particular thing. 
They then drop this qualification and assert that the learners 
are those who are ignorant as such, or those who do not know. 
If on the other hand he says that the learners are the foolish, the 
procedure is reversed: the sophists bring forward some parti- 
cular thing the learner does know and then go on to say that 
because he knows this thing, he is knowing or wise. Of course 

4 The fallacy is described by Aristotle {On Sophistical Refutations 1 66b 38 ff., 
trans. Forster) : 

Fallacies connected with the use of some particular expression absolutely or in a 
certain respect and not in its proper sense, occur when that which is predicated 
in part only is taken as though it was predicated absolutely (orav to ev /tepei 
Xeyofievov ws airXws elp-qptvov Xrj<f>dfj). For example, 'If that-which-is-not is an 
object of opinion, then that-which-is-not is'; for it is not the same thing 'to be 
something ' and ' to be ' absolutely. Or again, ' That-which-is is not, if it is not one 
of the things which are, e.g. if it is not a man'. For it is not the same thing 'not to 
be something' and 'not to be' absolutely; but owing to the similarity of the 
language, 'to be something' appears to differ only a little from 'to be', and 'not 
to be something ' from ' not to be '. 

Aristotle then goes on to describe the converse fallacy, a dicto simpliciter ad 
dictum secundum quid, e.g., an Indian is black but he has white teeth. There- 
fore it is (wrongly) argued that, because of his teeth, he is both white and 
not white. (In the Euihydemus there is an example of this fallacy at 299Bff.) 



in saying that the learners are the wise, Gleinias presumably 
thought himself to be asserting that people who are able to 
learn are intelligent, not that they are omniscient ; thus he was 
not rightly refuted by admitting there to be some special 
matter which the learners do not know. But since his remarks 
appeared to have led to the conclusion that the learners are the 
ignorant, and since wise and ignorant are clearly opposite, he 
believed himself to have been refuted when such was not in fact 
the case. Again, when he said the learners were foolish, he of 
course meant that they were ignorant of what they were about 
to learn, not that they were utterly stupid, so that again his 
refutation was not real but apparent. 5 

5 Mr. Richard Robinson, in an article, 'Plato's Consciousness of Fallacy', 
Mind, LI, No. 102, April, 1942, pp. 97-114, has given some attention to 
this passage, and comes to the (to me, at least) extraordinary conclusion 
that although Plato quite obviously regarded the Euthydemus arguments as 
fallacious (p. 102) this discussion of the equivocal nature of fiavddveiv 
gives us no reason to believe that Plato realized ' the peculiar subtlety and 
formidableness of this type of fallacy ' ; it contains nothing more than ' one 
of those crass ambiguities out of which puns are made' (p. 107). In fact, 
1 Plato appears to have remained till death at the point of view stated in the 
Euthydemus, that ambiguity [i.e. equivocation] is of no importance to the 
philosopher' (p. 1 14 — the concluding sentence of the article). As far as I can 
see, Robinson's chief reason for asserting that Plato attached no importance 
to the fallacy of equivocation is that there exists neither in the Euthydemus nor 
elsewhere in the dialogues any technical term (I suppose d/^tjSoAi'a or 
ofMcowfiia) for this fallacy. If there is no technical term for equivocation, 
nor, in fact, for fallacy in general, Plato can have had no consciousness of 
this fallacy nor of fallacy as such (pp. 102-3). And where there is no abstract 
consciousness of a thing, this thing cannot, apparently, have been thought 
important. (See his Plato's Earlier Dialectic, Oxford, 1953 (2nd ed.) pp. 4-5, 
for a general exposition of this view.) I would submit that since Plato is 
perhaps, of all philosophers, the least addicted to the use of technical 
terminology, the absence of such terminology proves nothing at all. One 
might as well argue that Plato attached no importance to metaphysics, 
since he has no name for this part of philosophy. In an earlier article in 
Mind, L, No. 98, April, 1941, pp. 140-155 on 'Ambiguity', Robinson also 
refers to this passage in the Euthydemus. Again he writes that Plato shows 
consciousness of ambiguity here, but goes on to say that at 278B such 
matters are dismissed as puerile. ' This profound contempt for the jugglers of 
language, and this serene confidence that their art can have nothing to do 
with the serious business of science, are dominant in Plato's early dialogues ' 
(p. 141). Plato's contempt for the jugglers of language is certainly profound, 
and it is certain that he refers to their tricks as 7rcuSia, but he does this, not 
because he thinks that their art has nothing to do with ' the serious business 


As we have seen, the verdict of Socrates on this sort of activity 
is that it is frivolity, and is so for the reason that it tells us 
nothing about the real state of things. This is a just accusation, 
since the sophists, in order to place Cleinias in the position of 
appearing to contradict himself at every turn, have been forced 
to change their own ground each time as well. Their contri- 
bution to the question they have raised concerning the relation 
between learning and wisdom (if we are charitable enough to 
suppose that they have really raised such a question) is exactly 
zero. They have used words as counters in a game, not as 
correct indices to things; hence they know nothing of the 
ovofxdrajv opdor-qg which Socrates ironically pretends that 
they are demonstrating. We may guess, too, that if virtue is 
thought to be knowledge, any art which produces no knowledge 
will obviously fail to lead to any kind of virtue — in fact it is not, 
properly speaking, an art at all. The art which does convey 
knowledge and which leads at least to the choice of virtue, is 
given a demonstration by Socrates in the next scene — a 
demonstration which could hardly contrast more sharply with 
the one which has just occurred. 

Scene II 278E-282D 

Do all men wish to prosper (ev irpdrr^iv) begins Socrates, 
with an apology for asking such a simple-minded question. 
This is obviously what all men wish, but how will they bring it 
about? Clearly by having many good things, such as wealth, 
health, physical beauty, noble birth, talent, and honor. In 
addition, men will wish to be temperate, just, and brave, and 
to possess wisdom. This seems to be a complete list of the goods 
men need in order to prosper, but one has been left out, namely, 
good fortune, which is really the same as wisdom. That these 
are in fact the same is illustrated by various examples to which 
Cleinias agrees: flute-players have the best luck in flute- 
playing, school-masters in writing, pilots in navigation, and so 

of science' but because (1) the consciousness of ambiguity which the sophists 
undoubtedly possess could be put to serious use, but in fact is not, and (2) 
their arguments stand in the way of the development of his own positive 
philosophy. That Plato conceived this fallacy and others as having a great 
deal to do with ' the serious business of science ' is one of the major contentions 
of this book. 



on. In every case, wisdom causes men to be fortunate (she can 
never make a mistake and still be wisdom) so that he who has 
wisdom has no need of good fortune as well. 

A man will be happy and prosper if he possesses many goods, 
continues Socrates 280B, but only if he receives some benefit 
from these goods. A thing cannot benefit us unless we use it, so 
that our happiness will depend on the use of our possessions. 
Furthermore, it will depend on their right use, since it can be 
shown that more evil results when a man uses a thing wrongly 
than when he leaves it strictly alone. As far as right use is 
concerned, nothing seems to bring this about except knowledge. 
It seems, then, that we cannot get any benefit from our other 
possessions unless we have this one. If a person lacks knowledge, 
it will be better to have and do little than to have and do much, 
since much action with no sense can only result in evil and 
misery. The whole question of goods, then, is not whether they 
are good in themselves but whether they are guided by wisdom 
or by ignorance. Other things are not in themselves good or bad, 
but wisdom is good and ignorance is bad. 

The conclusion must therefore be 282A that we should 
endeavor in every way to become wise, since wisdom is the 
means to the right use of things, and it is their right use that 
makes us happy. Cleinias agrees that he ought to be willing to 
perform any honorable service in order to become wise. Of 
course, says Socrates, wisdom may not be teachable, but may 
present itself to men of its own accord. For his part, Cleinias 
thinks it is teachable, thus saving Socrates a long inquiry into 
this point. If Cleinias thinks both that wisdom can be taught 
and that it is the only thing in the world that makes men happy, 
he surely cannot help saying that it must be pursued, and, in 
fact, intending to pursue it with all his might. This, says Cleinias, 
is exactly what he intends to do. 

In this second scene it is clear from the start that Socrates 
will be talking about 'how matters stand', not simply playing 
with words, since he follows up his first assertion, that all men 
wish to prosper, with the question, how are they to prosper. His 
discourse will lead Cleinias to the choice of something (the 
means to prosperity), whereas the demonstration by the 
sophists led to nothing but his verbal confutation. After this 
beginning, Socrates has only to show what he thinks the means 



to prosperity is (he thinks, of course, that it is wisdom), and 
Cleinias will not only say that it should be chosen, but will 
actually be disposed to choose it ; right speech will lead to right 

Socrates' first question is of especial interest in connection 
with right speech since it contains the equivocal expression 
ev irpdrreiv, a phrase which can mean either to fare well (to 
succeed) or to do well (to act rightly). However, Socrates 
makes no attempt to play on the ambiguity of the expression, 
as the sophists did in the case of ' learn ' in the previous scene. 
It is of course more than likely that Cleinias makes the choice of 
wisdom with the idea that he is choosing it as the means to a 
popular and not a philosophical happiness, and thus perhaps 
gets more than he bargained for, but the probability that he 
understands ev 7rpdrreLv in the sense of prosperity rather than 
virtue has no connection with the argument; the discussion 
could have proceeded in the same way if he had taken ev 
irpdrreiv in the other sense. 6 

Another indication of the fact that Socrates is concerned with 
the things denoted by words rather than purely with the words 
themselves is his freedom in the use of synonyms, evrvxeiv 
and evSaifjLovetv are substituted for ev TTpdrrew, for instance, 
and oo<j>La is replaced on occasion by iirurrqfirjy and even by 
<f>p6vr)(jis. Apparently it is a matter of indifference what terms 
are used so long as the general meaning is clear; as far as the 
argument is concerned, wisdom can equally well be called 
knowledge or prosperity happiness. With the sophists the 
situation is quite otherwise. Because they are arguing to a strict 
contradiction, and because they are doing so by means of 
devices which require the use of single expressions in different 
senses, they cannot use synonyms in this free fashion for fear of 
upsetting the argument. Socrates can do so, first, because the 
things to which he refers are clearly the same even if he expresses 
them differently, and, second, because his purpose is different : 
he is not arguing to a contradiction but exhorting to a choice. 
His discourse, when compared with that of the sophists, shows 

6 Cf. Shorey's note on Republic I 353E in the Loeb edition, 1953: '[The 
equivocation on ev TrpdrTeiv] does not seriously affect the validity of the 
argument, for it is used only as a rhetorical confirmation of the implication 
that K<xKd>s apx^v, etc. = misery and the reverse of happiness' (p. 105). 



us something of interest about the nature of correct speech : 
the form of language is less important than its ability to indicate 
correctly the things the speaker desires to express. 7 

In addition to what may be inferred from this scene about 
language, there are at least two other significant points which 
emerge from the discussion. The first is that the acquisition of 
wisdom is not represented as merely one among various ways of 
obtaining happiness, but as the only way. A man will be happy 
if he has many good things, but none of the things which are 
ordinarily considered good turn out to be so unless we possess 
this one thing, wisdom, which is good in itself. Without the 
guidance of wisdom, other so-called goods are indifferent in 
value ; if they are guided by ignorance, they may even be bad. 
Wisdom is at first included with all the other 'goods', but 
Socrates' pretense of having forgotten good fortune and then 
finding out that good fortune is after all the same as wisdom, 
has the effect of singling out wisdom for particular notice. 
When the two additional points are made, that a good is not 
a good unless it benefits us, and that nothing can benefit us 
unless we know how to use it, wisdom takes a definite place as 
the one intrinsically good thing among a group of others that 
are only derivatively good. 

The other notable point which results from the discussion is 
that wisdom is declared by Cleinias to be teachable at 282D. 
This assertion is not presented as the outcome of any logical 
progression, but appears to burst upon Cleinias as a spontaneous 
result of the process of exhortation to wisdom to which he has 
just submitted. (Cf. the 'leaping spark' of Ep. VII, 341C.) 
It will be remembered that Socrates, in his question to 
Dionysodorus at 274D, implied that eristic was incapable of 
justifying its own function; its practitioners claimed to teach 
virtue but could not show that virtue was in fact teachable. 
At the same time it was suggested that some other art could do 
this, this other art presumably being dialectic. A demonstration 
of dialectic has now taken place, and we see that Cleinias, who 
was simply confused and helpless at the end of the first demon- 

7 Cf. Protagoras 358B, where Socrates says, 'for whether you say pleasant 
or delightful or enjoyable, my excellent Prodicus, or in whatever style or 
manner you may be pleased to name these things, pray reply to the sense of 
my question. ' Cf. also Charmides 1 63D. 



stration, emerges from this one with the firm conviction that 
wisdom, if not virtue, can be taught. 

To say that wisdom can be taught is obviously not the same 
as to say that virtue can be, but the treatment of wisdom in 
what has just preceded suggests strongly that if wisdom is not 
the same as virtue, it is certainly very like it. Wisdom is the one 
sole good, for instance, as we see from 28 iE, and, when the 
future of Cleinias is spoken of, e.g., at 275A, 278D, 283A, it is 
said to be desirable not only that he should become wise, but 
that he should become virtuous and good. Wisdom and virtue 
thus appear to be linked in such a way that to aim at one is 
tantamount to aiming at the other. Whether the actual identity 
of the two in this dialogue be admitted or not, 8 the dialectical 
process has certainly yielded a result which is far more positive 
and striking than anything accomplished by eristic. 

Scene III 282D-288D 

Socrates now allows the sophists to continue their own demon- 
stration, but not until he has, innocently enough it seems, laid 
down for them the direction he wishes it to take. That is, he 
presents them with the alternative of showing whether Cleinias 
ought to acquire every sort of knowledge or only one ; in either 
case he assumes that they will want him to acquire some sort. 
Socrates continues his pretense of expecting to hear some 
wonderful arguments, and certainly gets them, in a way 
(davfiaGTov nva 283B). He again insists on his earnestness in 
the matter of the fate of Cleinias, but the sophists carry on in 
their previous vein; they do not exhort, but confute. 

The arguments which the sophists now employ are again 
fallacious and are again constructed by means of the two devices 
used in their first performance, the fallacy of equivocation and 
the fallacy of secundum quid. Since the brothers are this time 
playing on certain ambiguities in the verb 'to be' (i.e., in later 
terminology, they are deliberately confusing its use as copula 

8 E. S. Thompson, in writing of Euthydemus 282C in relation to the 
Meno (The Meno of Plato, edited with Introduction, Notes and Excursuses, 
London, 1901, p. 1), says 'The shifting of the subject of the question from 
apeT-q to oo<f>ta is not important ' ; he regards this whole passage as relevant 
to the question, is virtue teachable, as propounded in both the Meno and the 



with its existential use) , the results are much more serious from 
a philosophical point of view, as we shall see. 

Dionysodorus begins 283B by asking whether Socrates and 
his friends really and truly desire Cleinias to become wise. 
They assure him earnestly that this is the case. But if they wish 
him to become wise and not to be what he is now, i.e., ignorant, 
they evidently wish him to become what he is not, and no 
longer to be what he is. In other words, they wish him not to be, 
i.e., to be dead and gone. 

The mechanics of this first argument may be easily explained. 
Dionysodorus first obtains from Socrates the statement that he 
wishes Cleinias not to be ignorant. Then, by dropping the word 
'ignorant', he asserts that what Socrates has really said is that 
he wishes Cleinias not to be; i.e., not to exist. He has used the 
fallacy of secundum quid in moving from not being some particular 
thing (ignorant) to not being absolutely, and has at the same 
time altered the sense of ' be ', which when joined with ' ignorant ' 
was used as copula, but which, when taken alone, naturally 
suggests existence. 9 

The argument appears trivial enough but is actually of 
considerable significance, since it is based on a metaphysics 
incompatible with Plato's, namely, that of Parmenides. The 
essence of Parmenides' philosophy was, of course, that what is is, 
and what is not is not; consequently, there is no becoming. 
From the Parmenidean point of view, if Cleinias is at the 
moment ignorant, as his friends admit, then he must remain 
ignorant, since this is what he is. If his friends say they wish him 
to become something else, they are really talking nonsense since 
nothing else exists for him to become. Therefore they can only 
mean that they wish for his non-existence. The clearest solution 
of this difficulty occurs in the Sophist (e.g., 257Bff.) where 
Plato shows that it is possible for one thing not to be another 
thing and yet still exist ; it can be other instead of absolutely 
not. In this way he saves becoming and is consequently able to 
give some sort of reality to particulars. The Eleatic logic which 

9 On this passage see also E. H. Gifford, The Euthydemus of Plato with 
Revised Text, Introduction, Notes, and Indices, Oxford, 1905, who writes that at 
283D, ' the pronoun os is here equivocal, being used both in its proper sense 
as referring to a person and in an adjectival sense like ofos' (p. 36). This 
may be true, but the fallacy does not depend on this equivocation alone. 



is at the bottom of the majority of fallacies in this present scene 
is inextricably tied to a metaphysics which denies becoming; 
as a result it is necessary for Plato to expose these arguments, 
since in so doing he discredits the metaphysics as well. Until 
this work is done, he cannot provide a satisfactory basis for the 
theory of Forms. Here in the Euthydemus, the task is primarily 
that of exposure ; the constructive work is accomplished else- 
where. 10 

Ctesippus is very annoyed at the affront to his favourite 
Cleinias, and upbraids Dionysodorus for speaking so falsely 
283E. This gives Euthydemus the opening he needs for another 
fallacious argument, one which results in the denial of the 

10 E.g., Republic 476Dff. This passage in the Republic, which leads to the 
establishment of opinion as something midway between knowledge and 
ignorance 478D, is of exceptional interest, since the argument begins in a 
way which might well be the beginning of a sophistical argument. Socrates is 
considering what he and Glaucon might say to the man whom they have 
classified as having opinion, not knowledge. Socrates will ask questions and 
Glaucon is to answer on this man's behalf. The first question is (476E) 
' Does he who knows know something or nothing ? ' ' He knows something, ' 
Glaucon replies. 'Is it something that is or is not?' 'Something that is,' 
Glaucon answers, adding 'How could that which is not be known?' 
Socrates and Glaucon agree that what is is entirely knowable, and what in 
no way is is entirely unknowable. Up to this point the progress of the 
argument would be perfectly satisfactory to Euthydemus or Dionysodorus. 
These two would presumably proceed by bringing forward a case of some- 
thing which is not some other thing, and then, by use of secundum quid, would 
argue that this thing is absolutely not, and thus is absolutely unknowable. 
Socrates is of course in agreement with them that the absolutely not 
cannot be known (he says so at 47 7 A, as we have just seen) but he would 
disagree with them as to what things, if any, are to be placed in the category 
of the absolutely not. The sophists, who are working on the basis of a logic 
tied to Eleatic monism, would of course argue out of existence anything 
which is other than what is. (The thing which is may alter but is only one 
thing at a time.) Plato, who is concerned to recognize two orders of being, 
that of the Forms and that of the many participating in them (see above 
476A) diverges completely from the Eleatic type of argument at 477A 
when he has Socrates say ' If a thing, then, is so conditioned as both to be 
and not to be, would it not lie between that which absolutely and unquali- 
fiedly is and that which in no way is ? ' The Eleatic rule, of course, is that a 
thing must either be or not be, and there would, for a sophist of the Eleatic 
type, be no possibility of the existence of a thing which simultaneously is 
and is not. Plato has provided an object for his faculty of opinion in clear 
and conscious opposition to the Eleatic rule, and has, by implication, made 
a frontal attack on Parmenidean monism. 



possibility of false speech. He persuades Ctesippus to say that a 
person who lies speaks the thing about which he is speaking, or, 
in other words, speaks that which is. But to speak that which is, 
or to speak things that are, is to speak the truth, so Diony- 
sodorus cannot have been lying. Ctesippus objects, since who- 
ever spoke as Dionysodorus did does not say things that are. 
But, continues Euthydemus, the things that are not surely are 
not, and can be nowhere ; no one can possibly make them be in 
such a case. Now when orators speak in public do they do 
nothing? No, they do something, says Ctesippus. Then they 
also make, and speaking is a kind of making and doing. No one, 
therefore, speaks what is not, since he would then be making 
something, and Ctesippus has agreed that no one makes what is 
not. And if no one speaks what is not, no one speaks what is 
false; Dionysodorus therefore speaks what is true and what is. 
Ctesippus has to admit that Dionysodorus speaks what is, but 
he still protests that he does not speak it as it is. 

Here the argument relies again on the fallacy of equivocation. 
The verb 'to be ' (here t<x 6Wa) is again taken in two senses, as 
meaning either 'to exist' or 'to be true'. Conversely, its 
negation is taken to mean either 'not to exist' or 'to be false'. 11 
Therefore as soon as Ctesippus admits that the man who speaks 
speaks what is (by which he means only that the man speaks 
about a thing which exists) Euthydemus changes the sense of 
' to be ' and accuses him of having said that the man speaks the 
truth. When Ctesippus objects to this and says that Dionyso- 
dorus has spoken things that are not (by which he means things 
that are not the case, i.e., are false), Euthydemus again changes 
the sense of the verb and accuses him of having said that things 
that are not are (i.e., that what does not exist does exist). The 
sophist has, moreover, strengthened his position by insisting 
that speaking is a kind of making, by which he implies that 
speech produces some object. This enables him to suggest that 
while the man who speaks what is produces an object which 
exists, the man who speaks what is not is attempting to produce 
an object which does not exist. Since this is obviously im- 
possible, he is in a strong position to deny the possibility of 

11 Gifford, op. cit., p. 36, thinks, rather, that the equivocation is on 
Aeyeiv. But to bring off the complete argument it is necessary that to. ovra be 
equivocal as well. 

J 5 


speaking falsely. The real key to the matter, however, is his 
deliberate confusion of what is not true with what does not 
exist, a confusion which, when combined with the Eleatic denial 
of the real existence of what is not, leads irresistibly to the 
denial of falsehood. 12 

As a result of Ctesippus' continued insistence that Dionyso- 
dorus does not speak the things that are as they are {<l)s e^ei), 
we have next (284Cff.) a brief argument turning on the 
ambiguity in this phrase and also upon the ambiguity of kolkcos 
Xiyeiv, which can mean either to say that a thing is bad or to 
speak ill of this thing. The gist of this argument is that if there 
are some persons, e.g., good men, who speak of things as they 
are (i.e., tell the truth), then, since bad things are bad (or, in 
the Greek idiom, badly), they will speak badly of bad things, 
i.e., speak ill of them. In answer to this, Ctesippus says that 
Dionysodorus had best beware of being included among bad 
men, lest the good speak ill of him. Euthydemus supposes that 
the good also speak greatly of the great and hotly of the hot. 
And frigidly of the frigid, returns Ctesippus. Ctesippus is 
turning quite abusive, says Dionysodorus. No, says Ctesippus, 
he is merely giving the sophist a friendly warning to refrain 
from saying that he, Ctesippus, wishes his best friends to be 
dead and gone. 

At this point 285A, Socrates intervenes. If the two brothers 
really know how to destroy a wicked man and make him into a 
good one, then he and Ctesippus ought not to quarrel over a 
word (firj ovofiaTL hia<j>lp€adai). Socrates himself is willing to 
volunteer for such treatment ; the sophists may do anything they 
like to him so long as they make him good. Ctesippus agrees, 

12 It is perhaps not immediately obvious how the fallacy of equivocation 
could become part of the stock-in-trade of a neo-Eleatic sophist, since this 
fallacy appears to be based on the fact that a word has many meanings 
rather than one. (It would be natural to connect this fallacy with pluralism 
rather than with monism.) However, although the fallacy is based on the 
use of the same term in at least two different senses, its success as an instru- 
ment of refutation depends on keeping up the pretense that the term has 
only one sense. E.g., when Cleinias has asserted that it is the wise who learn 
and then has to admit that it is not after all the wise who learn but the 
foolish, he would not think himself refuted unless he supposed that ' learn ' 
had the same meaning throughout. The term is being used in two senses, 
as we have seen, and the solution of the fallacy consists in pointing this out. 



and tries to make his peace with Dionysodorus. After all, he 
was only contradicting him, and contradiction is not abuse. 

Socrates of course knows perfectly well that the sophists not 
only have no interest in making anyone virtuous, but could not 
do so even if they wished, since they have committed themselves 
to a metaphysics and a logic which is incompatible with change. 
I think that this denial of change or becoming on the part of the 
sophists explains why their style of argument can never succeed 
in showing that virtue or wisdom can be taught. Teaching is an 
activity which assumes that the pupil is capable of progressing 
from ignorance to knowledge, from vice to virtue, that he can, 
in fact, become other than what he now is. The eristic of 
Euthydemus and Dionysodorus makes no provision for teaching, 
learning, or any other kind of becoming, since it is so closely 
associated with the Eleatic logic that it remains confined to the 
static opposition between is and is not. If the sophists were really 
able, by some other means, to make a man good, their termin- 
ology would be of very minor importance, but the arguments 
they employ make it inconsistent for them to do any such thing. 

The argument against contradiction which occurs next 
(285Dff.) is cut to the familiar Eleatic pattern. Ctesippus has said 
that he is not abusing Dionysodorus but contradicting him, 
whereupon Dionysodorus affects surprise that Ctesippus 
should think there is such a thing as contradiction. His argu- 
ment against it is as follows : every thing has its own description 
(\6yos) which is as it is, not as it is not. No one speaks a thing 
as it is not, as they have already agreed. If both speak the 
description of the thing, they are speaking the same words and 
thus do not contradict; if neither speaks it, they do not deal with 
the thing at all, and thus do not contradict; if one speaks it and 
the other does not, only one of them gives the description and 
thus cannot be contradicted by the other, who gives no 
description. Again the combination of equivocation and 
secundum quid is at work: the man who speaks what is not, i.e., 
not what the other man speaks, is said not to speak at all. 
The qualification of 'is not' is dropped and the sense of the verb 
'to be' is altered from copulative to existential. 13 

13 With the argument against contradiction here in the Euthydemus cf. 
Sophist 236E : * Str. You see, Theaetetus, it is extremely difficult to understand 
how a man is to say or think that falsehood really exists and in saying this 

G 17 


Ctesippus is silenced by this latter argument, but not so 
Socrates. He has heard this sort of thing many times (the 
followers of Protagoras used it, as did others still earlier) and it 
always fills him with wonder since it seems to upset not only the 
views of others but itself as well. The substance of it seems to 
be that there is no false speaking ; if one speaks at all, one must 
speak the truth. This is apparently meant to extend to thinking 
also, so that there is no such thing as false opinion, nor is there 
ignorance, nor are there any ignorant men. Is it Dionysodorus' 
real view that there is no such thing as an ignorant man, or does 
he just want to say something startling? Dionysodorus replies 
that Socrates is to refute him. But how can there be such a thing 
as refutation if nobody ever lies ? There is no such thing, joins 
in Euthydemus, and Dionysodorus cannot have requested 
Socrates to refute him, since no one can request what is not. 
In response to this Socrates has what he says is a rather foolish 
question to ask : if there is no ignorance, there can never be any 
mistakes, and if there can be no mistakes, whatever in the 
world do the two sophists pretend to teach? Or did they not 
say that they could teach virtue to any man wishing to learn? 
How can Socrates be such an old simpleton as to bring up 
things that might have been said last year, says Dionysodorus, 
and yet be at a loss how to deal with what is being said at the 
moment. But, says Socrates, this phrase ' to be at a loss how to 
deal with it ' is just the same as ' to be at a loss how to refute it' — 
what else can the phrase mean? Dionysodorus refuses to 
answer, but insists on questioning Socrates. He asks whether 
things that 'mean' (voeTv) have life when they mean, or do 

not be involved in contradiction. Th. Why ? Str. The statement involves the 
bold assumption that not-being exists. But the great Parmenides, my boy, 
from the time when we were children to the end of his life, always protested 
against this . . . \ Plato's attitude towards Parmenides is always a mixed one, 
as may be seen, e.g., from a familiar passage at Theaetetus 183E (one not 
always quoted at sufficient length) : ' Parmenides seems to me to be, in 
Homer's words, "one to be venerated" and also "awful". For I met him 
when I was very young and he was very old, and he appeared to me to 
possess an absolutely noble depth of mind. So I am afraid we may not 
understand his words and may be still further from understanding what he 
meant by them; but my chief fear is that the question with which we 
started, about the nature of knowledge, may fail to be investigated, because 
of the disorderly crowd of arguments which will burst in upon us if we let them 
in . . . '. (My italics.) 



lifeless things ' mean ' also. Only the ones that have life, answers 
Socrates. Do you know any phrase that has life ? No. Then why 
ask what my phrase means? Socrates admits he has made a 
mistake, if mistakes are possible, that is. If he was not mistaken, 
they cannot refute him ; if he was mistaken, they were wrong in 
saying there is no such thing as a mistake. Thus their argument 
has not found out how to prevent itself from falling down in 
spite of the wonderful way it has of knocking down the argu- 
ments of others. 

The passage just summarized (286B-288A) contains 
Socrates' first really direct attack on the sophists. The exact 
nature of this attack is, however, rather puzzling, since although 
Socrates appears to say (286C, 288A) that the argument em- 
ployed is self-refuting, it is not altogether clear in just what way 
he regards the argument as involving its own refutation. The 
climax of his criticism (the 'clownish question' at 287A) 
emphasizes not the reflexive nature of the argument, but its 
inconsistency with the sophists' claim to be teachers. If Euthy- 
demus and Dionysodorus have argued ignorance out of exist- 
ence, they have obviously cut the ground from under their own 
professional feet, but to say that an argument which denies 
the possibility of ignorance is incompatible with the claim to 
teach is not at all the same as to say that this argument is self- 
refuting in its own nature. It is possible that Plato meant to 
imply that since the argument against refutation is itself based 
on a refutation (i.e., false speaking is argued against on the 
assumption that those who say that there is false speaking are 
themselves speaking falsely), therefore the argument is self- 
refuting. However, in default of any clear exposition of this 
point, it seems safer to assume that when Plato speaks of the 
argument as upsetting itself he has in mind simply the fact that 
the sophists, whose 'teaching' has turned out to consist of 
nothing but refutation, have made themselves look ridiculous 
by arguing that refutation does not in fact exist. From this 
point of view, the reference to Protagoras at 286G makes 
good sense, since that philosopher's assertion that man is the 
measure of all things exhibits a like incompatibility with the 
claim to teach. (Plato expounds this difficulty fully at Theaetetus 
161 Cff.) 

The positive result of Socrates' criticism is, therefore, to 



discredit the teaching ability of the sophists once again. This 
amounts, of course, to a renewal of his criticism of eristic. 
Socrates has made it clear before (e.g., 278B) that one of his 
objections to the sophistic method is that the sophists deal not in 
things but in words. This contrast is certainly implied at 286Dff., 
when Socrates asks, 'Is it merely to save your statement, 
Dionysodorus, that you state it so — -just to say something 
startling — or is it really and truly your view that there is no such 
thing as an ignorant man ? ' Dionysodorus' answer, ' But you . . . 
are to refute me', shows that he is not interested in discovering 
whether ignorant men do exist, but only in continuing the game 
of words. His attempt to continue it with the equivocation on 
' mean ' proves of course to be abortive since Socrates can now 
confront him with his inconsistency in accusing an opponent 
of making a mistake when he has just maintained that mistakes 
do not exist. The brothers are temporarily silenced, and 
Socrates, who is now top dog, can resume his ironic pretense 
that something fine will ultimately emerge from their discourse. 
Before permitting them another round, however, he resumes 
his own exhortation of Cleinias to wisdom and virtue. 

Scene IV 288D-293A 

At the end of their previous conversation 282D, Socrates and 
Cleinias agreed that wisdom ought to be pursued. This pursuit, 
called philosophy, is an acquiring of knowledge, and the know- 
ledge will be of a profitable kind involving a union of making 
and knowing how to use the thing made. Hence it will be unlike 
the arts of lyre-making, flute-making, or speech-writing. Even 
Cleinias can see that those who write speeches do not always 
know how to use them and vice versa. This latter art cannot be 
the one to make us happy in spite of its wonderful way of 
charming a crowd. Socrates suggests 290B that the art of 
generalship may be what they are looking for. Cleinias objects 
to this, since the general, who is a hunter of men, does not know 
what to do with his quarry; he hands it over to the politicians 
just as the huntsman and the geometer hand theirs to the caterer 
and to the dialectician. 

Crito is astonished that all this should have come from 
Cleinias, but Socrates is sure he heard this speech — it certainly 
was not spoken by Euthydemus or Dionysodorus. Socrates tells 



Crito that he and Cleinias went on examining various arts until 
they came to the kingly art (ff fiaoiXiKr) riyyt^ 29 iB). This one, 
which was the same as the art of the statesman, appeared to be 
the one to which all the other arts handed over their products 
to be rightly used. But, on considering the matter further, they 
found that while other ruling arts, such as medicine and agri- 
culture, had some effect to show as a result of their rule, this 
art apparently had none. Even if they defined it as the art by 
means of which men are made good, they were at a loss to say 
in what respect men were made good by this art. It should of 
course transmit some sort of knowledge, but knowledge of 
what ? Seemingly it can only give knowledge of itself. So Socrates 
and Cleinias found themselves no nearer than before to the 
discovery of what knowledge it is that makes men happy. In 
their distress they again called on the visitors to help them. 

Socrates' investigation has ended in a familiar paradox : the 
highest good has been identified with knowledge, but this 
knowledge has turned out to have no other subject than itself. 
(This is the same difficulty which occurs e.g., in the Charmides, 
in the case of temperance.) Socrates professes to be defeated at 
the end of the discussion, but it is obvious that he speaks 
ironically. He has, in fact, reason to be pleased with what has 
occurred, if only because his exhortation has resulted in the 
emergence of Cleinias as an independent thinker. (Again, of 
course, Plato means to exalt dialectic at the expense of eristic.) 
Furthermore, something has been accomplished on the philo- 
sophical side as well. 

In the earlier conversation between Socrates and Cleinias it 
became clear that wisdom, which was at first included in a class 
with other good things, was actually the only one among these 
goods which was good in itself; the others were good only when 
guided by wisdom. In other words, wisdom was lifted out of the 
class of things conventionally called good and given a special 
status of its own. The same is true of knowledge in the present 
passage. If I understand Plato rightly, he intends to say 
something like this: knowledge in the ordinary sense is know- 
ledge of something, it has some easily recognizable product. 
This product is good, but its goodness is dependent on the good- 
ness of knowledge as such. Knowledge as such has no identifiable 
subject-matter except itself simply because it comprises all 



* knowledges'. Socrates' discovery, therefore, that the kingly or 
political art possesses the special character of having only 
itself as subject-matter is, I think, meant to imply that it is 
identical with knowledge in itself and that it really is the art for 
which he and Cleinias have been seeking. It will meet the 
requirements they have agreed upon, since it will benefit men 
and make them happy. It will do this, however, not by teaching 
them any specific skill but by imparting an intellectual and 
moral quality which will enable them to rule over the specific 
skills and which will make these good by virtue of their subser- 
vience to the knowledge which is itself the only thing good in 

Thus there is a sense in which Socrates is no more a teacher than 
the sophists are, since neither of them is teaching a particular 
subject. The sophists pretend to teach by indulging in verbal 
tricks which have no connection with matters as they really 
stand. (They have, in addition, made the teaching endeavor 
inconsistent for themselves by the positions they have adopted 
on becoming and false speaking.) Socrates likewise is not really 
teaching wisdom and virtue, he is exhorting to them. That is, 
he is not attempting to convey a body of knowledge, but rather 
trying to convert the answerer to a state in which that person 
will be able to find out the body of knowledge for himself. 
The word 'convert 5 is accurate, since what Socrates is working 
towards is a change of soul. 

In the present scene, therefore, Socrates has not only con- 
firmed Cleinias in his intention of becoming wise, but he has 
given him some indication of the sort of wisdom he must acquire. 
Cleinias will no doubt fail to grasp the total implications of what 
has been said, but he has at least been set on the right road, since 
he is now convinced that wisdom must be pursued and that, 
when he finds it, it will be something useful and good. As a 
result he should be well-equipped to reject the spurious sub- 
stitutes which are being offered by Euthydemus and his kind. 

Scene F293A-304C 

Euthydemus makes a pretense of coming to the rescue of 
Socrates and Cleinias but it is immediately obvious that he is 
up to his old tricks when he asks 293B 'Would you rather, 
Socrates, that I instructed you as to this knowledge ... or 



propound that you have it?' As we are now accustomed to 
expect, he has no desire to talk about things but only about 
talking. The argument that follows again makes use of the 
fallacy of secundum quid. That is, from Socrates' admission that 
he knows one thing, Euthydemus moves to the conclusion that 
he is simply knowing (omniscient). Plato's mastery of this 
fallacy is complete, as may be seen from the way in which 
Socrates makes it difficult for Euthydemus to bring the 
refutation off. Socrates does this by adding qualifications to his 
answers, qualifications intended to prevent the sophist from 
moving from the particular to the general. For instance, when 
Euthydemus asks at 293B, 'Do you know anything?', Socrates 
replies, 'Yes . . . many things', and again, when Euthydemus 
says, 'Then you are knowing if you really know', Socrates 
answers, 'Certainly, in just that something' (tovtov ye avrov). 
Socrates continues 293G to prevent Euthydemus from placing 
him in the category of either the knowing or the not knowing by 
insisting that while there are some things he knows, there are 
also many he does not know. When Euthydemus tries the 
opposite tack and wants to say that Socrates is absolutely not 
knowing, Socrates answers, 'Not in that thing' (eKelvov). 
Euthydemus then invokes the law of contradiction to show that 
Socrates cannot claim both to be and not be the same man 
(i.e. to be and not to be the man who knows) in regard to the 
same matter at the same time; since this is an impossible 
position, he must admit that he has been shown to know every- 
thing, including that knowledge he was seeking. Quite aside 
from the fact that the sophists have argued against contradiction 
earlier, this is wrong on two counts in the present context : ( 1 ) 
Socrates has not really allowed himself to be placed in either 
category, and (2) he has not said he did or did not know the 
same matters, but different ones. Socrates, however, does not 
trouble to make either of these objections. He contents himself 
with trying to place the sophists in the same predicament in 
which they have placed him, i.e., of being both knowing and 
not knowing. 'Tell me', says Socrates at 293E, 'you both know 
some existent things, of course, and others you do not?' 
Euthydemus and Dionysodorus will have none of this, since 
such an admission would destroy their position. Instead, they 
claim to know everything, as in fact do the rest of mankind, 



since c they cannot know some things and not others, and so be 
at once knowing and unknowing' (294A). Socrates and 
Ctesippus press them to display their knowledge of some 
particular point, but this they decline to do, having no desire to 
go from the universal to the particular; they merely continue to 
repeat that they know everything. 

As Socrates still pretends incredulity, Euthydemus embarks 
on a slightly different form of the same argument. Socrates, 
however, continues to qualify his answers as before, much 
to the sophist's annoyance. For instance, when Euthydemus 
asks whether Socrates knows 'by one and the same means 
always ... or sometimes by one and sometimes by another' 
(296A), Socrates answers, 'Always, whenever I know, it is by 
this means '. Socrates is afraid, he says, that ' this word "always " 
may bring us to grief. This is of course the crucial point, since 
as soon as Euthydemus can connect 'always' with 'know', he 
can set up his argument as before. This he proceeds to do, more 
or less over Socrates' dead body. He decapitates the phrase 
'you know always by the same means' to 'you know always' 
(296C). Socrates lets this pass for a moment because he has an 
extremely damaging conclusion to draw from the premise with 
which he is confronted. He shows 296E that if he knows all 
things, then he must know not only true statements, such as 
'good men are just', but also their (false) opposites, e.g., 'good 
men are unjust'. Dionysodorus, who is a little less wary than 
his brother, falls into the trap and asserts that Socrates cannot 
know this statement. He is reprimanded by Euthydemus, who 
sees that if Socrates can prove that there is one thing he does 
not know, they will lose their advantage. Socrates, meanwhile, 
is quick to add to their discomfiture by pointing out that it would 
be impossible for Dionysodorus to speak wrongly if he really 
knew everything. 

The sophists wriggle out of this latter difficulty by picking up 
the word 'brother', which Socrates has dropped by the way, 
and are presently back at the old stand, happily engaged in 
confounding that which is other with that which is not. By 
means of this fallacy and by the use of equivocations on such 
words as 'his', 'yours', and 'mine', they establish in turn that 
Socrates has no father, that any man who is a father is father of 
all, and that Ctesippus' father is a dog. Then follows a brief 



argument 299B to the effect that if anything, e.g., medicine, or 
arms, or gold, is a good, then one ought to desire as much of it 
as possible in all times and in all places. (This is a dicto simpliciter 
ad dictum secundum quid; what is good absolutely is said to be good 
in some particular ridiculous way.) The mention of a talent of 
gold in one's skull 299E leads Ctesippus to pay the sophists back 
in their own coin by a play upon the ambiguity in the word 
' theirs ' : the skulls belong to the Scythians, therefore when they 
drink out of them, they drink out of their own skulls. Euthy- 
demus, taking advantage of the fact that Ctesippus has spoken 
of the Scythians as gazing at their skulls, begins in return to 
exploit the ambiguity in Swara opdv 300A, which can (by 
reason of the Greek accusative and infinitive construction) be 
twisted to mean either 'possible to see' or 'able to see', and then 
in similar fashion to work with oiytovra Xeyetv, which he takes 
at will as either a silent person speaking or as a speaking of 
something silent. Dionysodorus again shows himself to be some- 
what less on his guard than Euthydemus by answering ' Neither 
and both' to Ctesippus' question 300D 'are all things silent or 
do they speak ? ' Ctesippus is now sufficiently familiar with the 
sophists' methods to seize upon the word ' both ' as a slip, since 
of course the eristic tricks are largely dependent on the classi- 
fication of everything into one of two exclusive classes. (Socrates, 
for instance, was said to be either knowing or not knowing, 
but not both.) 

The brief passage 300E-301 C which now follows is of unusual 
interest since it is the only place in the dialogue which appears to 
refer to the theory of Forms. 

From the Eleatic point of view, the chief objection to the 
theory must be that it calls for two distinct sorts of existence, 
that of the Forms, and that of the particulars participating in 
them. Dionysodorus' question to Socrates at 300E, whether he 
finds beautiful things to be different from or the same as the 
beautiful, appears to be put with this objection in mind. If 
Socrates should answer that beautiful things are the same as the 
beautiful, Dionysodorus could argue that there is no distinction 
between them; if Socrates should answer that they are different, 
he could argue, by the usual fallacious equation of the other 
with the not, that beautiful things do not exist. In either case, 
monism would be preserved. Socrates, apparently perceiving 



the direction of the sophist's attack, attempts to answer the 
question in the terms in which it has been asked and yet to 
avoid representing the relationship between particulars and 
Forms as one of either total sameness or total difference: he says 
that beautiful things are different from the beautiful but that 
each has some beauty present with it (krepa . . . avrov ye rod 
kolAov ' Trdpeart pbivroi eKaarcp avrcov /caAAos" rt 30 1 A). 

The notion of 'presence with' is an obstacle to the type of 
refutation the sophist presumably has in mind, since it provides 
some sort of mediation between beauty and beautiful things 
without, however, making them identical. He therefore attacks 
this notion in his next question, saying ' So if an ox is present 
with you, you are an ox, and since I am present with you, you 
are Dionysodorus '. This is interesting on two counts, first that 
Dionysodorus has taken presence in the purely physical sense, 
and second that he has destroyed the contrast between Forms 
and particulars by dealing in terms of particulars only. The two 
points are interrelated and are both connected with monism. 
Dionysodorus, in order to confine himself to one sort of 
existence, has spoken not of oxness as present with Socrates (as he 
ought to do if he is to meet the theory of Forms on its own 
ground) but of an ox as so present. Since both Socrates and the 
ox are physical particulars, the only possible mode of presence 
is physical presence. Dionysodorus has made Socrates' answer 
look ridiculous, but he has been able to do this only by dis- 
torting the nature and sense of Socrates' terminology. 

Dionysodorus now follows up the joke about the ox by phras- 
ing what is basically the same objection in another and more 
subtle way. He says 301 A 'In what way can one thing, [i.e. the 
different] by having a different thing present with it, be itself 
different?' (^4AA<x riva rpoirov . . . eripov erepco Trapayevo- 
[livov to erepov erepov av etr];) In this new question he 
is still attacking the idea of 'presence with', but is at the 
same time taking advantage of Socrates' statement at 301 A 
that beautiful things are different from the beautiful itself. His 
procedure is to disregard what it is that both the beautiful and 
beautiful things are different/rom (each other) and to refer to both 
as simply 'different', or, if it suits his purposes better, as 'the 
different'. This is the familiar fallacy of secundum quid at work, 
and the result of the manoeuvre is again, as in the case of 



Socrates and the ox, to destroy the distinction between Forms 
and particulars. We can see how this has happened by putting 
down the sophist's question with its correct expansions. It 
should read, 'In what way, because of beauty being present 
with beautiful things, can the beautiful things be beautiful?' 
In its truncated state, however, it has become, in effect, ' How 
can the different be different ? ' The distinction between beauty 
and beautiful things has been swallowed up, and with it the 
distinction between two sorts of existence. Furthermore, the 
question is now no longer one concerning the mode of partici- 
pation of particulars in their homonymous Form, but one which 
inquires about the possibility of a Form being predicated of 
itself. The Form about which this question is asked is one which 
the sophist has constructed by the transformation of the 
adjective 'different' into an adjectival noun, 'the different'. 
This 'Form' he apparently assumes to have the same status 
as the beautiful, so that he raises no objection when Socrates 
argues that if the beautiful is beautiful and the ugly ugly, then, 
by analogy, the different will be different and the same the 
same. Socrates has given a strong hint that he is refuting the 
sophist by fallacious means since he says at 30 iB, 'already I was 
attempting to imitate the cleverness of these men, I was so eager 
to get it'. We have reason, therefore to question whether he 
regards the statement that the beautiful is beautiful as a 
legitimate one. It seems to me likely that he does not, since this 
statement is so clearly modelled on the sophist's truncated 
question about the different. 14 However, the legitimacy of this 
statement, though an interesting topic in its own right, 15 is not 

14 With this passage cf. Sophist 254Dff. and A. L. Peck, 'Plato and the 
MEriETA rENH of the Sophist', Classical Quarterly, 1952, pp. 46-9. 

15 For a discussion of the self-predication of Forms in Plato, see G. Vlastos, 
'The Third Man Argument in the Parmenides\ Philosophical Review, LXIII, 
1954, pp. 319-49. (See also articles subsequent to this paper, a bibliography 
of which may be assembled by consulting R. S. Bluck, 'Forms as Standards', 
Phronesis, 2, 2, 1957, pp. 115-27, and N. B. Booth, ' Assumptions involved in 
the "third man" argument', Phronesis, 3, 2, 1958, pp. 146-9. To this list 
should also be added the excellent paper of R. E. Allen, ' Participation and 
Predication in Plato's Middle Dialogues', Philosophical Review, LXIX, i960, 
pp. 147-64.) A number of scholars, of whom T. A. deLaguna is perhaps the 
first ('Notes on the Theory of Ideas', Philosophical Review, XLIII, 1934, 
p. 450), have also noticed that an instance of the self-predication of Forms 



the main point at issue. The point is that Socrates himself has 
used an analogy between this statement and the illegitimately 
constructed statement about ' the different ' to bring about what 
is not a refutation of the sophist's original objection to Socrates' 
account of the relation between Forms and particulars, but a 
refutation of the distorted shape in which this account now 
appears, and he has done this by means of exactly the same 

occurs at Protagoras 330G. (See Vlastos, op. cit., p. 337, who calls this 'the 
star instance of Self-Predication in Plato'.) I feel myself that Socrates, when 
he asks Protagoras at 330C, 'Is justice a thing or not' (1} hinaioavvq 
wpay/xa ri ianv tj ov8ev 7rpay/ia), is deliberately misleading the sophist, 
and that he asks the question in order to persuade Protagoras to place 
justice in the same class as particular just acts. When Protagoras does this, 
it then appears reasonable, as it would not perhaps otherwise have done, 
to inquire whether this thing, justice, is itself just or unjust. This whole 
passage needs to be studied in the general context of the reduction of the 
various virtues to knowledge, and of the ultimate refutation of Protagoras, 
who has attempted to maintain both that the virtues are distinct 330A and 
that virtue can be taught (317B, 319B). What Socrates does in the dialogue 
is to show that Protagoras' claim to teach virtue implies not, what he has 
also asserted, that the virtues are distinct, i.e., many, but that they are one, 
i.e., that all can be reduced to knowledge. The argument which employs 
the statement 'justice is just' is the first step in the reduction of the virtues. 
By this means, and by means of some additional fallacies such as the passage 
from 'not just' to 'unjust' (331A), two of the virtues, justice and holiness, 
are equated. Socrates then proceeds to equate two others, temperance and 
wisdom. This he does by a clearly fallacious argument : the term 'folly' is used 
in one sense as opposite to temperance and in another sense as opposite to 
wisdom — the term is employed equivocally, in other words. Socrates is 
about to bring together the two complexes justice-holiness and wisdom- 
temperance by a third argument (which will apparently make use of an 
equivocation on 'temperate') when Protagoras, presumably detecting the 
direction of the argument, interrupts with his speech on the relative 
profitableness of things at 334Aff. The main lines of the argument are not 
picked up again until after the Simonides interval, i.e., not until 349B. The 
remainder of the dialogue is then devoted to the more difficult task of the 
reduction of courage to knowledge. This argument also appears to be 
fallacious, and consciously so, but it is too long to be analysed here. (See 
G. Vlastos' Introduction to Plato's Protagoras in the Library of Liberal Arts 
series, New York, 1956, p. xxxi. Professor Vlastos would not agree with me 
that the fallacies are deliberate, however.) The fact then that the argument 
employing the statement 'justice is just' appears as one step in a series of, as 
I think, consciously fallacious arguments, leads me to suspect strongly that 
this argument is consciously fallacious also. Furthermore the present passage 
in the Euthydemus strikes me as confirmatory evidence. 



To sum up : Dionysodorus has attacked the theory of Forms 
by means of his accustomed weapons of equivocation and 
secundum quid, and has, as a result, completely misinterpreted 
the theory because the contrast between Forms and particulars 
has disappeared. Socrates has defended himself against the 
attack, not by pointing out the fallacies involved (as he does, for 
instance, in the case of the equivocation on 'learn' in the first 
scene), but by accepting the misinterpretation of the theory 
and then proceeding to refute the sophist in the sophist's own 
terms. He has made him pay the penalty for constructing an 
illegitimate Form (the different) by forcing him to accept an 
analogy between it and legitimate Forms like the beautiful and 
the ugly. In a sense the theory, in its correct version, has not 
been defended at all, but Socrates has at least demonstrated 
that this sort of fallacious attack can be disposed of on its own 
grounds. He seems to me also to have demonstrated the complete 
incompatibility of the theory with Eleatic monism in a very 
striking way, since the fallacies which the sophist employs are 
Eleatic in origin and their effect upon the theory of Forms is to 
destroy its essential dualism. 16 

Of additional interest in connection with this passage is the 
fact that it is here that the sophists have come nearest to being 
philosophers. The relationship between Forms and particulars 
is undoubtedly the most difficult aspect of the theory of Forms 
and, in selecting this for attack, they show an instinct for a 
genuine philosophical problem. Then the ox joke, although 
based on a misinterpretation of the theory, does at least draw 
attention to the fact that one meaning of 'present with' is 
physical presence, and that Socrates has not taken this into 
account. Furthermore, by implying that in some way the 
presence of the ox with Socrates causes Socrates to become an 
ox, they have indirectly raised the question whether the beauty 
of beautiful things is caused by the presence of beauty with 
them. This is again a genuine philosophical problem, and one 
not touched upon by Socrates. Even the fallaciously constructed 

16 If Dr. A. L. Peck is right in his interpretation of the first part of the 
Parmenides (I think he is) this same incompatibility is demonstrated even 
more strikingly in that dialogue. See 'Plato's Parmenides: Some Suggestions 
for its Interpretation', Classical Quarterly, N.S., vol. Ill, Nos. 3 and 4, July- 
October, 1953, pp. 126-50. 



question, how can the different be different, involves two 
interesting philosophical points: can a Form be made from 
a relational adjective, and can a Form be predicated of itself? 
I think that Plato, in this passage, means to point out the very 
thin partition between philosophy and sophistry: the sophists 
have raised some important philosophical issues, but their 
love of refutation and word-play is so strong that they are 
incapable of investigating any of them seriously. 

The argument about cutting up the cook which follows 30 1 D 
is not of great interest, since it is merely another instance of a 
fallacy of the same sort as those involving the phrases Swara 
6pdv and oiywvra Aeyetv (i.e. it is again made possible by the 
Greek accusative and infinitive construction). This is followed 
30 1 E by another argument based on the equivocal use of 
'mine' and 'yours' which places Socrates in the position of 
claiming to sacrifice his ancestral gods 303A. ' Bravo, Hercules 
... a fine argument, ' exclaims Gtesippus. ' Now do you mean 
that Hercules is a bravo, or that bravo is Hercules ? ' asks Diony- 
sodorus. At this there is wild applause from all sides. Even 
Socrates himself joins in praise of the clever pair who have 
acquired such a great accomplishment in such a short time. 
Obviously they care nothing for the multitude, but only for 
those of their own sort, since most people would consider it a 
greater disgrace to refute with such arguments than to be 
refuted by them. Also they show very good manners in being 
willing to stitch up their own mouths as well as those of others. 17 
Best of all is the way in which anyone else may learn this skill 
so quickly, as Ctesippus has been observed to do. However, 

17 The passage 303D reads ' when you say there is nothing either beautiful, 
or good, or white, and so on, and no difference of things at all, in truth 
you simply stitch up men's mouths' and so on. The sophists have not 
explicitly maintained these propositions in the dialogue, so that it is difficult 
to see to what Socrates refers. My guess is that this is an oblique reference 
to the fact that the sophists, in their attack on the theory of Forms 301 ff., 
have, because of their monism, denied the existence of beautiful things. Thus 
they must, to be consistent, deny the existence of all other sorts of particulars, 
good things, white things, etc. Their monism of course leads them also to the 
denial of the existence of the difference of things, since if any one thing is, 
any other thing, differing from it, does not exist. The denial of false speaking 
earlier in the dialogue can also be regarded as a denial of the difference of 
things, since it destroys the distinction between truth and falsity. (See below 
on the Cratylus pp. 48 ff.) 



they will be wise not to use it in too many public discussions in 
case other people should learn the whole thing straight away 
and give them no credit for it. They should talk only to each 
other, or, if anyone else is present, be sure to charge him a good 
fee. (With this speech of Socrates, the related portion of the 
dialogue ends.) 

Epilogue 304C-end 

Socrates again suggests to Crito that they should both go to 
school to the sophists. Crito is willing, but fears he is one of those 
who would prefer to be refuted by such arguments rather than 
to use them himself. He wants to tell Socrates of a conversation 
he has just had with a man who heard the discussion and who, 
as a result, was very critical not only of philosophy but also of 
Socrates, for being willing to take part in such a performance. 
Socrates suspects that this was one of those clever persons, 
half-way between philosophy and politics, who write speeches 
for the law-courts. This man feels that Euthydemus and the 
like stand in his way, and thus seeks to decry philosophy in 
order to enhance his own reputation. He ought to realize that a 
person midway between two good things is inferior to both of 
them, and that he really occupies the third place, not the first. 
Still we should be grateful for anyone who says anything bor- 
dering on good sense. Crito is much in doubt what to do about 
his sons, when he sees the sort of persons who profess to educate 
people. But, says Socrates, he should not be deterred from 
philosophy simply because there are some bad teachers in the 
business. He should test it himself, and if he finds it to be what 
Socrates maintains it is, then he should pursue it, he and his 
children too. 

To decide whether Isocrates or some other is referred to in 
this last section seems to me not of great importance in so far 
as the general meaning of the passage is concerned. Since 
Socrates, in being 'so strangely willing to lend himself to 
persons who care not a straw what they say, but merely fasten 
on any phrase that turns up ' (305A) , does not seem, to the casual 
observer, to differ greatly from the sophists, it is time for a 
defence of philosophy, both as an occupation and as a method 
of instruction. The criticism made by the semi-philosopher 
semi-politician serves to provide the occasion for this defence. 



The defence itself, however, is curiously indirect. It begins 
by exposing the inconsistency in the speech-writer's position. 
According to the principle laid down by Socrates at 3o6Aff. 
this man can only be right in regarding himself as superior to 
both the philosopher and the politician if philosophy and politics 
are both bad. If both are good, he is inferior to both; if one is 
good and the other bad, he is worse than one, but better than 
the other. It is obviously to his advantage to malign both pro- 
fessions, but, according to Socrates at 306G, he is apparently 
unwilling to do this. Thus, we are intended to conclude, his 
attack is inconsistent. 

Since Crito, at 305 A, has just quoted the speech-writer as 
saying that both philosophy and the people who follow it are 
worthless and ridiculous, it seems rather strange that Socrates 
should represent the man as declining to admit that either philo- 
sophy or politics is bad. The only possible explanation of this 
seems to be that Socrates is trying to bring out the fact that the 
speech-writer, must, in his own mind, be making some distinc- 
tion between a bad philosophy practised by the sophists and 
some other pursuit, also called philosophy, which he regards as 
good. This explanation is borne out by what occurs at the very 
end of the dialogue 307Aff. since here Socrates advises Crito 
not to eschew philosophy because some, or even most, of its 
practitioners are worthless. (We can think of two he has in 
mind.) What Socrates has been trying to achieve in the rather 
roundabout preliminaries to this final point is a separation 
between the genuine philosophical activity and its false imita- 
tion. This he does by bringing forward the case of a man who 
attacks philosophy and yet is unwilling to admit that philosophy 
is bad. The inference, which is left for us and for Crito to draw, 
is that what was being attacked under the name of philosophy 
was really eristic. Philosophy in its proper sense remains a good, 
and as such, is unaffected by the practises of those who pretend 
to engage in it but are in fact doing something else. Crito must 
of course test the matter for himself and make the decision 
about his own education and that of his sons accordingly, but 
we are certainly left with the impression that he can hardly fail 
to decide in favor of philosophy as Socrates conceives it. The 
alternating demonstrations of dialectic and eristic have had 
their effect upon him as well as upon Cleinias, and the dis- 



cussion has accomplished its purpose even at second-hand. 
It is Plato's obvious intention that it should accomplish its 
purpose with the reader as well. 

The foregoing summary of the Euthydemus has shown that the 
results of the dialogue are chiefly two : the recommendation of 
philosophy conceived as dialectic (since it is by this means that 
an individual may be exhorted to the choice of wisdom and 
virtue), and the rejection of philosophy conceived as eristic. 
Plato's main reason for rejecting eristic is that it disregards 
correct speech and thus gives no reliable information about the 
true state of things. 18 Euthydemus and Dionysodorus have 
treated words as counters in a game; hence their verbal tricks 
can convey nothing as to how matters really stand, and, as 
Plato demonstrates rather than asserts, can lead to no alteration 
in the individual soul. The rejection of eristic is made especially 
crucial for Plato by two circumstances: first, that the sophists 
propose to teach virtue by this means, and second, that their 
arguments are for the most part based on a metaphysics (that 
of Parmenides) which is incompatible with his own. In exposing 
the fallacious nature of these arguments, therefore, Plato is not 
only defending his own method of teaching but also clearing the 
way for what is, after all, the indispensable condition of this 
method, the theory of Forms. 

I now propose to consider, in the next two chapters, some 
fallacious arguments similar to arguments in the Euthydemus. 
As already stated in the introduction, I believe this similarity to 
warrant the assumption that their employment by Plato in these 
other dialogues is as deliberate as it clearly is in the Euthydemus. 
The philosophical issues involved will again be of considerable 
significance, as we shall see. 

18 This does not mean, contrary to what is said in the Cratylus, that Plato 
wants to say here that correct speech leads to knowledge of the nature of 
reality. What he does want to say, I think, is that we should employ the 
best terms we have, without perverting their accepted meanings. 



Theaetetus i63Aff. 

very early in the dialogue, at 145E, Socrates asks the young 
Theaetetus for a definition of knowledge. Like various other 
persons questioned by Socrates, Theaetetus at first fails to grasp 
the nature of general definition, but he soon comes to under- 
stand what Socrates wants, and, at 15 iE, he puts forward the 
view that knowledge is perception. Socrates' reaction to this is 
to point out that such a definition is tied both to the relativism 
of Protagoras and to the all-flowing philosophy of the Hera- 
cliteans. He proceeds to develop a number of consequences of 
the flowing philosophy, but does so in such a way that 
Theaetetus is not clear as to whether this philosophy (and thus 
his own definition) is being attacked or not 157G. Even after the 
passage i57Eff. in which the flowing philosophy turns out to 
provide no criterion by which waking may be distinguished 
from dreams or sanity from insanity, Theaetetus is apparently 
still in doubt, since he raises no objection when Socrates says 

Therefore you were quite right in saying that knowledge is nothing 
else than perception, and there is complete identity between the 
doctrine of Homer and Heraclitus and all their followers — that all 
things are in motion, like streams — the doctrine of the great philo- 
sopher Protagoras that man is the measure of all things — and the 
doctrine of Theaetetus that, since these things are true, perception 
is knowledge. 

But after the speech 161 Gff. in which Socrates suggests that 
Protagoras' measure might as well be a pig or a dog-faced 



baboon, Theaetetus' faith in his definition is severely shaken; 
he says 162C, 'when we were discussing the meaning of the 
doctrine that whatever appears to each one really is to him, I 
thought it was good; but now it has suddenly changed to the 
opposite.' It is at this point, when Theaetetus has begun to 
wonder whether his definition will hold after all, that the pas- 
sage with which I am concerned in this chapter begins. 
'Well then,' Socrates says at 163A, 

let us look at it in this way, raising the question whether knowledge 
is, after all, the same as perception, or different. For that is the 
object of all our discussion, and it was to answer that question that 
we stirred up all these strange doctrines, was it not? 

He then proceeds to give a set of three arguments, all of which 
show that the identity of knowledge with perception can lead to 
absurdity. The first two arguments play upon the ambiguity 
in the verb 'to know' (emo-ra^at) in a manner similar to 
that in which the arguments in the first scene of the Euthydemus 
played upon the ambiguity in the verb 'to learn' (fxavddveiv) , 19 
The third argument employs the 'knowing — not knowing' 
dichotomy which occurred in the Euthydemus in Scene V as the 
basis of the sophists' claim to be omniscient, and which is also 
implied in the arguments of the first scene, as Socrates observes 
at 278A. 

The first argument (which has two parts) takes ' know ' in the 
sense of 'perceive' and also in the sense of 'learn'. It begins by 
using the example of a person hearing a foreign language he 
has not yet learned. If knowledge is really the same as percep- 
tion, then when this person hears the language, he does in fact 
know it, or else, if he does not know it, he does not in fact hear 
it. Both conclusions fail to fit the facts as originally described. 
The same argument is then repeated for seeing : if we see letters 
we have not yet learned, we either know them after all because 
we see them, or we do not really see them because we do not 
know them. In the case of both the foreign language and the 
letters, an example has been given of a kind of knowing which is 
obviously not co-extensive with perceiving. The point is, of 

19 'Know' and 'learn' were terms which were notoriously productive of 
sophistical arguments. Cf. Theaetetus 1 99A, 077-77 ns x a ^P €L ^Xkwv to iiriaTaadai 
Kai navOdvew. 



course, that although the word ' know ' is not entirely unmean- 
ingful in connection with hearing or sight, there is at least 
one other sense of ' know ' with which the idea of perception is 
incompatible, and if these two meanings are combined, it is 
possible to construct contradictions. The two sets which have 
been given may be readily schematized as follows : 

A (i) I hear German = (3) I perceive German = (5) I know 
(2) I have not learned German = (4) I do not know German = 
(6) I do not perceive German. 

B (1) I see Greek letters = (3) I perceive Greek letters = (5) I know 
Greek letters. 
(2) I have not learned Greek letters = (4) I do not know Greek 
letters = (6) I do not perceive Greek letters. 

(The contradictions are double, i.e., between (3) and (6), and 
(4) and (5), in each case.) 

Theaetetus' answer at 163B, that 'we know just so much of 
[the language and the letters] as we hear or see . . . [but we do 
not know] what the grammarians and interpreters teach about 
them', shows that he understands perfectly that it is the double 
use of 'know' which has produced the contradictions. 

The second argument shows that there is still another sense of 
' know ' which is not accounted for if knowledge is said to mean 
simply perception, i.e., the sense of remembering what has been 
learned. The question is asked at 163E whether a person 
remembers, i.e., knows, when his eyes are shut, i.e., when he is 
not perceiving. Theaetetus answers that it would be absurd to 
deny such a thing, so that we again have a case in which the 
identity of knowledge and perception has resulted in a paradox: 

G (1) I remember x = (3) I know x = (5) I perceive x. 

D (2) I do not see x = (4) I do not perceive x = (6) I do not know x. 

(The contradictions are between (3) and (6), and (4) and (5), 
as before.) 

In the first argument, it turned out to be possible to perceive 
without knowing ; now it turns out to be possible to know with- 
out perceiving. Knowledge and perception cannot therefore be 
said to be the same, but we must say that they are different 



This conclusion is not allowed to stand for more than a 
moment, however, since Socrates immediately objects 164C, D, 
that he and Theaetetus have been 'acting like professional 
debaters' (avnXoyiKws) and have 'based [their] agreements 
on the mere similarity of words ' (ras rcov ovofxarajv o/xoAoyi'a?) ; 
they ' do not see that [they] who claim to be, not contestants for 
a prize (aytovioTaL) , but lovers of wisdom (fiiXoaofoi) are 
doing just what those ingenious persons do' (rots- beivols 
avSpdaw) . Furthermore, if Protagoras had been alive, he would 
not have allowed the definition to be treated in such a fashion. 
Socrates, therefore, undertakes to come to Protagoras' assis- 
tance, 'for a man might find himself involved in still worse 
inconsistencies than those in which we found ourselves j ust now, 
if he did not pay attention to the terms which we generally use 
in assent and denial' (165A). We then have a third argument 
against the definition, which is also eristic, 20 and which 
Socrates describes as 'the most frightfully difficult question of 
all' (165B). 

Socrates begins by asking Theaetetus whether the same 
person can both know and not know that which he knows. 
This, Theaetetus says, is obviously impossible. But suppose 
someone were to cover one of your eyes with his hand, and then 
to ask if you saw his cloak with that eye ? Theaetetus answers 
that he would say he did not see the cloak with that eye but did 
see it with the other. Then, says Socrates, your opponent would 
conclude that you both see and do not see at the same time, and 
since, by the definition, seeing is the same as knowing, that you 
also both know and do not know at the same time. Theaetetus 
objects that his seeing and not seeing was only 'after a fashion' 
(ovtoj ye TTOis 165C), but this qualification would not be 
allowed, according to Socrates. Furthermore he goes on to show 
(i65Dff.) that the identity of knowledge and perception could 
give rise to a host of similar difficulties, any one of which, if 

20 It might be well to point out that I have usually employed the terms 
'eristic', 'sophistical' and so forth in relation to arguments similar in form 
to those used by Euthydemus and Dionysodorus even when such arguments 
are not used contentiously, since, for my purposes, it was necessary to find 
some convenient way of indicating the similarity between arguments in the 
Euthydemus and arguments elsewhere. (Incidentally I argue below, pp. 43- 
45, that the fact that Socrates employed sophistical arguments in the 
Theaetetus does not in the least imply that he did so in a sophistical spirit.) 



employed by 'a nimble fighter, fighting for pay in the war of 
words', would be sufficient to reduce Theaetetus to a state of 

Since Socrates has come to the aid of Protagoras and his 
definition by demonstrating that still another paradox may be 
generated by means of the very definition Protagoras would 
like to defend, we may reasonably regard this as a rather odd 
type of assistance. His offer is ironic, as we may have guessed 
from his remark to Theodorus at 165A: 'Now see how I shall 
help him' (aKeipai ovv rfy y'ifirjv fiorjOeiav). Nevertheless, the 
very perverseness of his procedure has the effect of bringing 
into prominence the objections which he attributes to Protagoras 
following the argument at i66Aff. The substance of all these 
objections is that perception has been conceived in a way which 
is too static. 

Do you suppose you could get anybody to admit that the memory 
a man has of a past feeling he no longer feels is anything like the 
feeling at the time when he was feeling it ? Far from it. Or that he 
would refuse to admit that it is possible for one and the same person 
to know and not to know one and the same thing ? Or if he were 
afraid to admit this, would he ever admit that a person who has 
become unlike is the same as before he became unlike? In fact, if 
we are to be on our guard against verbal entanglements, would he 
admit that a person is one at all, and not many, who become 
infinite in number, if the process of becoming different continues ? 

What Plato seems to be having Protagoras point out is that it is 
neglect of becoming which gives rise to fallacious arguments of 
the type just used. Plato means, I think, to imply that the 
Heracliteanism with which he has associated Protagoras gives 
at least a hint of the correct way of avoiding such fallacies, 
although a completely flowing philosophy would err too much 
in the opposite direction. (This last he has already suggested in 
the passage i58Eff. in which Socrates ill turns out to be wholly 
other than Socrates well.) The fallacies which have been set up 
have all been constructed on an Eleatic basis (e.g., it is assumed 
that 'know' has only one meaning, whereas it in fact has 
several) and this fact is thrown into relief by their occurrence in 
a context of Heracliteanism. The first set of arguments displays 
the fact that the definition of knowledge as perception leaves 
out of account the phenomenon of learning; in the second 



argument we see that it also fails to account for memory. The 
third argument, in which knowing and not knowing are 
regarded as exhaustive alternatives, neglects the possibility of 
any mediation between the t wo extrem es, or, in more Platonic 
language, it fails to distinguish the other from the not. Pro- 
tagoras' objections have of course served to draw attention to 
the fact that these various forms of becoming have been 
neglected, but they do not, nevertheless, suggest the correct 
solutions; for these we must go to the comments given to 

In both parts of the first argument, Theaetetus appeared to 
be contravening the sophistic rule that it is impossible for the 
same person both to know and not know the same thing at the 
same time. He escaped from this difficulty by the distinctions 
made at 163B, the implications of which were that it is perfectly 
possible to know in the sense of perceiving and at the same time 
not know in the sense of not having learned. In the second argu- 
ment, Theaetetus again has to go against the sophistic rule 
about knowing and not knowing, but the explanation is again 
that ' know ' has been used in two senses : it is possible to know 
in the sense of remembering what one has learned and at the 
same time not to know in the sense of not at the moment 
perceiving this thing. Socrates' remarks about people who 
behave avriXoyiKOJs are directed towards the way in which the 
arguments have been set up, and towards the fact that 
Theaetetus has met them by asserting, in effect, that it is after all 
possible to know and not know at the same time. Theaetetus 
has of course taken this position with the understanding that 
'know' is an equivocal term, but Plato wishes to demonstrate 
that a sophist would ignore this: i.e., he would use the equi- 
vocation, but would conceal the fact. Socrates therefore asks 
Theaetetus outright whether it is possible for a person both to 
know and not know that which he knows, and to the question 
put in this bald form, Theaetetus of course says no. Plato then 
employs the technique already illustrated from the Euthydemus 
of dropping the qualifying part of a phrase (the fallacy of 
secundum" quid) . That is, "although Theaetetus wishes to say 
(correctly) ' I do not see with the eye which is covered, but I do 
see with the eye which is uncovered', Socrates points out that 
his sophistical opponent would interpret this as an assertion of 



seeing and not seeing. Since seeing has been previously identified 
with knowing, Theaetetus is placed in the position of having 
contradicted his original statement that simultaneous knowing 
and not knowing is impossible. 

The third argument thus serves a purpose slightly different 
from that of the first two. In (i) and (2), we have seen that 
Plato is not only pointing out that the identification of know- 
ledge with perception may give rise to sophistry because ' know ' 
is an equivocal term, but also that the definition does not account 
for two important psychological facts: learning and memory. 
(Since both are matters of becoming and degree, they are 
bound to clash with the sophists' rule of either-or.) The third 
argument has the more limited purpose of showing that 
'knowing or not knowing' is itself a source of sophistry, but only 
so when put in its extreme form. Theaetetus has been using it 
correctly in (1) and (2) ; when it is formulated in the sophists' 
way, in (3), he quite rightly denies its validity. 

In connection with this interpretation of the three arguments, 
it is important to take note of some additional passages later in 
the dialogue. 

At 186E, after Socrates has made the distinction between 
sensations and reasoning, the definition of knowledge as per- 
ception is finally given up altogether. Theaetetus then offers 
a new definition, that knowledge is true opinion (rj dXr]dr)s 
S6£a 187B). He has not, he says, defined it as simply opinion, 
since there exists false opinion as well. Socrates immediately 
undertakes to examine, not the new definition, but Theaetetus' 
statement about the existence of false opinion. The way in 
which he begins the examination is very instructive: 

This then, at any rate, is possible for us, is it not, regarding all things 
collectively and each thing separately, either to know or not to 
know them? For learning and forgetting, as intermediate stages, I 
leave out of account for the present, for just now they have no 
bearing upon our argument. 

Th. Certainly, Socrates, nothing is left in any particular case except 
knowing or not knowing it. (i88Aff.) 

As soon as the 'knowing — not knowing' dichotomy is made 
operative, in other words, all forms of becoming must be 
dropped. This was so in the case of the sophistical arguments 



which Socrates brought against the 'knowledge is perception' 
definition, and is so again here when Socrates prepares to bring 
a similar type of attack against the existence of false opinion. 

The attack against false opinion consists of two parts. The 
first is based on the assumption just quoted, that to know an 
object implies knowing it completely and infallibly and not to 
know an object implies knowing nothing about it whatsoever. 
(There is no room for any intervening degrees of knowledge, in 
other words.) On this assumption there can be no explanation 
of false opinion since, as Plato shows by the method of 
exhaustion, such opinion cannot arise through a man think- 
ing either (i) that things he knows are other things he knows, 
or (2) that things he does not know are other things he does not 
know, or (3) that things he does know are things he does not 
know, or (4) that things he does not know are things he knows 

The second part of the attack is introduced by Socrates' 
suggestion that he and Theaetetus should substitute ' being and 
not being' for 'knowing and not knowing'. This of course paves 
the way for an argument of the type which the sophists 
employed against false speaking and contradiction in the 
Euthydemus 283Eff. The similarity shows clearly at the end of the 
argument 189A: 

Soc. So, then, does not he who holds an opinion hold an opinion 

of some one thing? 
Th. He must do so. 
Soc. And does not he who holds an opinion of some one thing hold 

an opinion of something that is ? 
Th. I agree. 
Soc. Then he who holds an opinion of what is not holds an opinion 

of nothing. 
Th. Evidently. 
Soc. Well then, he who holds an opinion of nothing, holds no 

opinion at all. 
Th. That is plain, apparently. 
Soc. Then it is impossible to hold an opinion of that which is not, 

either in relation to things that are, or independently of them. 
Th. Evidently. 

However, although Socrates has gone through two eristic 
arguments, he does not draw the typical eristic conclusion. 



That is, instead of saying that false opinion does not exist, he 
concludes that it 'is something different from holding an 
opinion of that which is not ' ( 1 89B) . Plato is still determined to 
show that there is such a thing as false opinion (the resulting 
absurdities would otherwise be too numerous 1 90E and cf. Sophist 
24 1 D, E) and he has rehearsed these arguments to show that 
they are not really effective against it. This we have had a 
reason to expect in advance, since we have been warned ahead 
of time that becoming will be left out. After the rejection of 
another possible argument against false opinion (which takes 
the form of a demonstration that it cannot be aAAoSo£ta or 
interchanged opinion), Socrates changes the direction of the 
discussion : he stops his negative defense of false opinion and 
shifts to a positive defense, that is, to the images of the wax 
tablet and the aviary. It is at this point that learning 191 G 
and presently memory 191D are reintroduced. It therefore 
seems clear that in this part of the dialogue as in the passage 
i63ff. there is a direct connection between the presence of 
sophistry (of a particular type) and the absence of becoming. 21 
(What I have loosely called becoming includes two things in the 
present context : ( 1 ) the psychological phenomena of learning, 
forgetting, and memory, and (2) degrees of knowing and not 
knowing.) This is presumably what Socrates has in mind when 
he says 157B that 'anyone who by his mode of speech makes 
things stand still is easily refuted'. But, as has been said earlier, 
a situation of total becoming would be, if not the producer of 
similar sophistical arguments, at least the producer of similar 
results, since, as Socrates has pointed out at 1 70G, the doctrine 
that man is the measure also abolishes false opinion. The 
reasons are different but the outcome is the same. 22 

21 Euthydemus and Dionysodorus would never have defined learning as 
'growing wiser about that which one learns', as Socrates does at 145D, for 
instance. It is interesting to note also that at Sophist 259G a criticism of the 
kind of quibble which results in the 'complete separation of each thing from 
all' precedes the introduction of the eiScDv ovuttXokt} at 259E. In order to 
get rid of Eleatic sophistry, in other words, some device must be employed 
(here Kowwvia rather than the introduction of becoming) in order to 
avoid the extreme type of separation which is ' the utterly final obliteration 
of all discourse ' (259E). 

22 By the same token we find that in the Cratylus the two oppposed theories 
of language (one Eleatic and one Heraclitean) really come to the same thing 



The arguments in the Theaetetus 1636°. are, it seems clear to 
me, very similar to arguments in the Euthydemus, and thus I 
have conducted the analysis of them on the assumption that 
Plato was completely conscious of their fallacious character. 
His method of introducing such arguments is, however, not the 
same as his method in the Euthydemus. In that dialogue, the 
fallacies appeared in the mouth of the two eristic sophists; 
where explanations were given, either directly or indirectly, 
they appeared in the mouth of Socrates. In the Theaetetus, 
however, it is Socrates who puts forward the fallacies; the 
explanations are, for the most part, given by Theaetetus. The 
shift in method is of course occasioned by the difference in 
Plato's attitude towards the respondents of Socrates in the two 
dialogues. Euthydemus and Dionysodorus, as purveyors of a 
spurious training in virtue, are the legitimate objects of ridicule. 
Theaetetus, on the other hand, is probably the most promising 
example of the philosophic mind to be found in the dialogues. 
Plato's treatment of him therefore differs toto caelo from his 
treatment of the sophists, a fact which comes out particularly 
clearly in Socrates' speech to Theaetetus at i54Dff. : 

Well, if you and I were clever and wise and had found out every- 
thing about the mind, we should henceforth spend the rest of our 
time testing each other out of the fulness of our wisdom, rushing 
together like sophists in a sophistical combat, battering each other's 
arguments with counter arguments. But, as it is, since we are ordinary 
people, we shall wish in the first place to look into the real essence 
of our thoughts and see whether they harmonize with one another 
or not at all. 

The real difficulty, however, is not to explain why Plato has 
given Theaetetus a role similar to that which was previously 
played by Socrates (we are probably right to suppose that this 
is his way of paying the young student a compliment) , but to 
explain why he has given Socrates a role similar to that which was 
previously played by Euthydemus and Dionysodorus. But the 
difficulty is only a difficulty if we assume that because Socrates 

in the end. Cf. below pp. 50-52. In the Theaetetus we may note another 
close connection between the philosophy of rest and the philosophy of 
motion at i58Eff. where we find that Socrates ill turns out to be 'wholly 
other' (i<ofii8fj tripos) than Socrates in health. The doctrine of total flux has 
resulted in an Eleatic use (or rather misuse) of trepos. 



is represented as employing eristic arguments, he is represented 
as employing them in an eristic spirit. If, for example, we 
compare Socrates' employment of fallacy in the Theaetetus 
passage with its employment by Euthydemus and Dionysodorus 
in the first scene with Cleinias, it will readily be seen that 
although the actual arguments are similar, the way in which 
they are used is quite different: 

( i ) Socrates is engaged in the testing of a specific definition, 
knowledge is perception. The sophists, however, have no 
definite philosophical task in hand; they are doing no more 
than attempting to substantiate their vague claim to be teachers 
of virtue. 

(2) Socrates wants Theaetetus to grow 163C, whereas the 
sophists have no interest in Cleinias except to make him look 

(3) Socrates shows clearly that the arguments presented are not 
his own arguments but are those of contestants {aycjviarai) 
who behave avnXoy lkws i 64C, or of ' a nimble fighter, fighting for 
pay in the war of words ' {jTeXraarLKos avrjp pnodo^opos iv Xoyois 
ipofievos 165D). In the Euthydemus, however, there is no doubt 
but that the sophists are intended to be completely identified 
with their own arguments ; they are in fact precisely the ' nimble 
fighters' whom Socrates has in mind. (Cf. eV rots Xoyois 
fidx^adai 2 72 A.) 

(4) Socrates handles his arguments in such a way as to bring 
out the important facts that (a) although there is some connec- 
tion between knowledge and perception, the complete identifi- 
cation of the two would restrict the meaning of ' knowledge ' in 
such a way as not to allow for two additional meanings, i.e., 
learning and memory, and (b) that the ' knowing — not knowing ' 
dichotomy is at the basis of a certain group of paradoxes which 
can be produced by use of the identification in its extreme form. 
The sophists in the Euthydemus are using similar arguments but 
with no thought of exposing the mechanics of these arguments — 
quite the contrary in fact. It is left for Socrates to point out that 
ILavdaveiv sometimes means 'learn' and sometimes means 
Understand' (278A) ; if this equivocation were pointed out by 
Euthydemus or Dionysodorus, the paradoxes they have just 
set up would be destroyed. 



All in all, we may conclude that Socrates' use of fallacy in 
this section of the Theaetetus is as clearly constructive as its use 
by the sophists in the Euthydemus is destructive. In a sense, of 
course, Socrates has used fallacy destructively, i.e., to destroy 
the complete equation of knowledge with perception. But he has 
performed this destructive task in such a way as to show that the 
definition of knowledge will be incomplete unless it accounts for 
the positive facts of learning and memory. Furthermore, by the 
rejection of the definition of knowledge as perception, he has 
paved the way for the two definitions which Theaetetus will 
offer later on: that knowledge is S6£a dX-qdrjs 187B, and that it 
is So£a aXrjdrjs /xera Xoyov 20 1 D, each of which is an advance 
over the definition which has gone before. It is clear, then, 
that fallacy has played an important part in the progress of 
the dialogue in the direction of a satisfactory definition of 

In the next chapter we shall again encounter fallacy con- 
sciously employed by Plato, but shall find it occurring not in the 
mouth of Socrates but in that of his opponent Cratylus. 
In offering a solution to the fallacy, Plato will again be doing 
constructive work, since he will be lending support to the theory 
of Forms. 



Cratylus 42gBfF. 

at 42 7D, Socrates has finished the long list of etymologies 
which, at the request of Hermogenes, he has produced to 
illustrate the theory of the natural correctness of names. 
According to Hermogenes, this theory is the one held by Craty- 
lus, but Cratylus himself has not yet entered the conversation. 
Now, however, Socrates asks Cratylus for his assistance in 
reconsidering the theory, since he, Socrates, has begun to dis- 
believe in his own wisdom 42 8D. Cratylus readily agrees with 
Socrates that names are given for the purpose of instruction, and 
that this instruction is an art whose artisans are the law-givers, as 
Socrates has said before at 388E. But when Socrates wishes to 
show, by an analogy between naming and painting, that some 
law-givers do their work better than others, Cratylus objects; he 
will not concede that one name is better and another worse. He 
maintains, on the contrary, that all names are correct, all that 
are really names, that is. Further questioning by Socrates 
reveals the source of Cratylus' reluctance to admit the existence 
of names which are false or incorrect : 

Soc. How about the name of our friend Hermogenes, which was 
mentioned a while ago? Shall we say that it is not his name at 
all, unless he belongs to the race of Hermes, or that it is his name, 
but is incorrect ? 

Cra. I think, Socrates, that it is not his name at all ; it appears to be 
his, but is really the name of some one else who possesses the 
nature that makes the name clear. 

Soc. And when anyone says that our friend is Hermogenes, is he not 



even speaking falsely? For perhaps it is not even possible to say 

that he is Hermogenes, if he is not. 
Cra. What do you mean? 
Soc. Do you mean to say that it is impossible to speak falsehood at 

all ? For there are, my dear Cratylus, many who do so, and who 

have done so in the past. 
Cra. Why, Socrates, how could anyone who says that which he says, 

say that which is not ? Is not falsehood saying that which is not ? 
Soc. Your reasoning is too clever for me at my age, my friend. 

However, tell me this : Do you think it possible to speak falsehood, 

but not to say it ? 
Cra. Neither to speak nor to say it. 
Soc. Nor utter it or use it as a form of address ? For instance, if some 

one should meet you in hospitable fashion, should grasp your 

hand and say, 'Well met, my friend from Athens, son of Smicrion, 

Hermogenes', would he be saying or speaking or uttering or 

addressing these words not to you, but to Hermogenes — or to 

nobody ? 
Cra. I think, Socrates, the man would be producing sounds without 

Soc. Even that reply is welcome ; for I can ask whether the words he 

produced would be true, or false, or partly true and partly false. 

Even that would suffice. 
Cra. I should say that the man in such a case was merely making a 

noise, going through purposeless motions, as if he were beating a 

bronze pot. (429B-430A.) 

The basis of the position taken by Cratylus is easily recogniz- 
able as the Eleatic argument against false speaking with which 
we are already familiar from the Euthydemus 286C. Cratylus' 
train of thought is apparently as follows : if a name is not the 
name of the object or person to which it is applied (as, for 
instance, in the case of Hermogenes, whose names does not 
express his real nature) then that name is not, or does not exist. 
Therefore a name either names correctly or is not a name at all. 23 
Again as in the Euthydemus, the argument depends on the use of 

23 It is possible to see a resemblance between the position taken here by 
Cratylus about names and that taken by Thrasymachus at Republic I 
34oCfF. to the effect that a ruler who makes a mistake is not really a ruler. 
Plato, in opposition to both Parmenides and Heraclitus (and Protagoras), is 
always concerned to allow for error: see, e.g., Theaetetus 190E 'if [false 
opinion] is found not to exist, we shall be forced to admit many 
absurdities . . . '. 



two fallacies: equivocation and secundum quid. (A shift in the 
meaning of 'is 5 from copulative to existential occurs in con- 
junction with a shift from the statement that a name is not the 
name of some particular person or thing, to the statement that 
it is not absolutely.) 

After some further discussion, Cratylus is persuaded to admit 
that there may perhaps be incorrect names, but his conversion 
is only temporary and he suffers a number of relapses before the 
dialogue ends (e.g. 432 A, 433C, 438C). The argument against 
false speaking has influenced him to such an extent that he 
cannot give it up even when Socrates has demonstrated in the 
clearest possible way that it is inconsistent to maintain both that 
all names are correct, and, what Cratylus wishes to say also, 
that the relation of names to things is one of representation or 
likeness (ojiiouo/za) . The view that all names are correct is one 
which Cratylus seems to have arrived at by means of the 
assumption that a name cannot be in any degree unlike the 
thing named, or, in other words, it cannot be in any way not 
that thing. But if a name is to be the likeness or image of a thing, 
it must in some fashion be unlike or not the thing, else it would 
be an exact copy of that thing. (Socrates illustrates this latter 
point by the example of the two Cratyluses at 43 2 Bff.) Therefore 
if Cratylus wishes to retain a theory of naming by represen- 
tation and likeness (a theory which he much prefers to naming 
by chance signs 434A), he will need to give up the view that 
there are no degrees of correctness in names. And he will also, 
by implication, need to give up the fallacious argument against 
false speaking which is associated with it. 24 

Now it appears from various passages in the Cratylus that 
Plato is extremely anxious to defend the view that there are 

24 At Sophist 24oBff. we see again that there is incompatibility between 
Eleaticism and the view that images or likenesses have some real status: 

Str. That which is like, then, you say does not really exist, if you say it is not true. 
Th. But it does exist, in a way. (A\X eon ye [vt\v ttu>s.) 
Str. But not truly, you mean. 
Th. No, except that it really is a likeness. 

Str. Then what we call a likeness, though not really existing, really does exist ? 
Th. Not-being does seem to have got into some such entanglement with being, 
and it is very absurd. 

A little further on at 241 D there is a passage giving a good illustration of 
the general clash between Eleaticism and the possibility of false speech : the 



both true and false names. 25 If, for instance, we go back to the 
beginning of the dialogue, we find Socrates questioning 
Hermogenes as follows: 

Soc. ... Is there anything which you call speaking the truth and 

speaking falsehood ? 
Her. Yes. 

Soc. Then there would be true speech and false speech. 
Her. Certainly. 
Soc. Then that speech which says things as they are is true, and that 

which says them as they are not is false. (385!^.) 

From this beginning Socrates then proceeds to argue that the 
parts of true speech are also true, and the parts of false speech 
false. Names are the smallest parts of speech so that it is possible 
to utter both true and false names. 26 After establishing this con- 

Eleatic Stranger tells Theaetetus that he will have to test the theory of his 
father Parmenides since 

unless these statements [that after a fashion not being is and on the other hand in 
a sense being is not] are either disproved or accepted, no one who speaks about 
false words or false opinion — whether images or likenesses or imitations or 
appearances — or about the arts which have to do with them, can ever help being 
forced to contradict himself and make himself ridiculous. 

25 In the early part of the dialogue, e.g. 385C, Plato usually speaks of 
names as 'true' (dX-qdes) or 'false' (ifjevSos), whereas later he tends to say 
' correct ' (6p66s) and ' incorrect ' (ovk 6p66s) . But the two sets of terms are 
clearly interchangeable when applied to names (cf. 430D), so that I have 
felt free to vary the usage in what follows. 

26 As it stands, this argument looks like an example of the fallacy of 
division, that is, it takes distributively two attributes, truth and falsity, 
which should be taken only collectively. (See Richard Robinson, 'A Criti- 
cism of Plato's Cratylus\ Philosophical Review, LXV, 1956, p. 328.) On the 
other hand, the premisses needed to establish the conclusion are both 
present in the text : 

385B13, C14 (A) The parts of true and false speech are true and false. 
C9 (B) Names are parts of true [and false] speech. 
C16 .'. (C) Names are true and false. 

The fact that most people would not accept (A), either because they hold 
the Aristotelian view (de Interpretatione i6ai3ff.) or because at Sophist 262-264 
Plato himself seems to attach truth and falsity to sentences rather than to 
names (nouns and verbs) , is irrelevant to the structure of the argument here 
in the Cratylus. (Incidentally, I think myself that in the Sophist passage there 
is nothing to indicate that Plato, in attributing truth and falsity to sentences, 
necessarily implies that they should not be attributed to names, but this is 
another matter.) I think, however, that Robinson is right, when he says, loc. 
cit. 3 that Plato commits the converse fallacy, that of composition, at 43 1 B. 

e 49 


elusion, his next move is to embark on a refutation of Hermo- 
genes 5 view that names are right by convention and agreement. 
This he does by showing Hermogenes that the convention 
theory, if carried to its logical conclusion, implies the relativism 
of Protagoras. Protagoras' theory, that man is the measure of 
all things, would involve Hermogenes in two admissions which 
he is unwilling to make: first, that no men are better than 
others, and, second, that no men are wiser than others 386B, C. 
The convention theory would involve Hermogenes also in a 
species of universal predication theory, attributed by Socrates 
to Euthydemus ('that all things belong equally to all men at the 
same time and perpetually') which has similar results ('for on 
this assumption also some could not be good and others bad if 
virtue and its opposite were always equally possessed by all' 
386D) . The point is that the theory of Protagoras and the theory 
of Euthydemus succeed equally in destroying intellectual and 
moral distinctions. If these distinctions are destroyed, then no 
man is better than another and name-giving cannot be a 
matter for experts, since no experts exist. And name-giving 
involves no particular wisdom, if all names are equally true. 
With these implications apparently in mind, Plato argues, on 
the contrary, that 'the giving of names can hardly be as 
[Hermogenes] imagine [s] ... a task for trifling or casual 
persons' 390D; names are made by the law-giver with the 
advice of the dialectician. Since Hermogenes can think of no 
answer to this, Socrates concludes that 

Cratylus is right in saying that names belong to things by nature 
and that not every one is an artisan of names, but only he who keeps 
in view the name which belongs by nature to each particular thing 
and is able to embody its forms in the letters and syllables (390 D, E). 

This is not in fact exactly what Cratylus is reported to have said 
nor what he will say later on, but the present point at issue is 
that Hermogenes is confronted with the failure of his theory 
(the convention theory) and that, furthermore, this failure is 
connected with Socrates' assertion that there are both true and 
false names. In other words, the convention theory is associated 
with the view that all names are true, whereas the natural 
rightness theory (in its present guise as a theory involving expert 
knowledge of the nature of the things named) is associated with 



the view that there are false names as well. 27 The two theories 
are obviously incompatible, as incompatible, one might say, 
as are the theories of Protagoras and Euthydemus with the 
theory of Forms. (The convention theory implies both of the 
former, as we have seen, and the natural Tightness theory, as 
it is here developed in connection with the notion of the ideal 
shuttle, certainly appears to imply the latter.) It seems, there- 
fore, as if the existence of false names were an essential element 
not only in the defeat of the convention theory but in the estab- 
lishment of the theory which Plato wishes to substitute for it, 
that is, the theory that the law-giver gives names with his eye 
on the ideal name, and that his work is judged by the dialec- 
tician. This giving and judging of names is obviously work for 
experts and requires the existence of both wisdom and wise men. 
And if wisdom exists, so does ignorance, and it is possible for 
names to be more or less well given. The well-given names are 
true and the others false. The false names are none the less 
names, but they, presumably, have been devised in ignorance 
of, or without attending to, the ideal names. 

This entire passage is of great interest, since it shows, in a 
fairly small compass, the essential points of Plato's quarrel both 
with the Heracliteans and with the Eleatics. Protagoras, who 
here, as in the Theaetetus, is grouped with those who favor a 
universal flux, is wrong, since on his view all men are equally 
wise and good. Euthydemus, who is here as much of an Eleatic 
monist as he is in his own dialogue (the view attributed to him 
at Cratylus 386D is a fair corollary of the doctrine of universal 
knowing which he and Dionysodorus maintain at Euthydemus 
294ff.), is also wrong, since on his view the conclusion is the 
same : all men are equally wise and good. In the case of Prota- 
goras, the conclusion comes about because each man is assumed 
to be the judge of his own shifting perceptions; in the case of 
Euthydemus, the conclusion is the result of a sophistic trick 
which argues, e.g., from knowledge of some particular thing to 
knowledge absolute. The two situations are totally different, 

27 This distinction between two versions of the theory of natural Tightness 
(one associated with the theory of Forms and the other with the Eleatic 
position to be taken up by Cratylus 42gBff.) is one not normally made but is, 
I believe, present in the text and is an important element in my general 
interpretation of the dialogue. See below pp. 52-55. 



but the outcome is the same — no man is better than another 
and no man is wiser than another. Wisdom and virtue (and 
their opposites) have, in effect, ceased to exist. 

That this is a state of affairs which would be intolerable to 
Plato need hardly be said ; the theory of Forms is inextricably 
tied to the view that there can be a progress from ignorance to 
knowledge, and hence that there must be allowance made for 
the possibility of error. It is also associated with the existence 
of an intellectual and moral aristocracy. Since the Cratylus is 
concerned with language, Plato has expressed his dissatisfaction 
with the two opposed positions by arguing not for the existence 
of folly or of error but for that of false names, but his intention to 
contrast the theory of Forms with both Heracliteanism and 
Eleaticism seems to me none the less clear. This contrast is 
concisely expressed at 386D when Socrates, after ascertaining 
that Hermogenes is unwilling to accept the views of either 
Protagoras or Euthydemus, formulates the view which is, in his 
mind, the necessary alternative to these: 

Soc. Then if neither all things belong equally to all men at the same 
time and perpetually [the position of Euthydemus] nor each 
thing to each man individually [the position of Protagoras], it is 
clear that things have some fixed reality of their own, not in 
relation to us nor caused by us; they do not vary, swaying one 
way and another in accordance with our fancy, but exist of them- 
selves in relation to their own reality imposed by nature. 

When Hermogenes has agreed to this, Socrates then goes on to 
develop the analogy between the ideal shuttle and the ideal 

Up to the point at which Socrates actually begins to illustrate 
the natural Tightness theory, then, we have some reason to 
associate this theory with the theory of Forms. This is parti- 
cularly clear in the passage 390D, E quoted earlier: Hermogenes 
and the convention theory have been refuted (partly by means 
of the analogy with Protagoras and Euthydemus, and partly 
by means of the passage about the ideal names) and Socrates 
concludes as a result that 

Cratylus is right in saying that names belong to things by nature, 
and that not every one is an artisan of names, but only he who keeps 
in view the name which belongs by nature to each particular thing 
and is able to embody its forms in the letters and syllables. 



Before Socrates begins to illustrate the theory, however, we are 
warned that it will probably not prove satisfactory. Socrates 
tells Hermogenes 39 iB that the best way to investigate the 
theory will be with the help of those who know. Hermogenes 
could, for instance, ask his brother Callias to tell him what he 
has learned from Protagoras on this topic — a suggestion Hermo- 
genes naturally rejects since he has already declared himself to 
be out of sympathy with Protagoras' views 386A. Or he could 
learn about the correctness of names from Homer. This sugges- 
tion Hermogenes for the moment accepts, but we can guess 
that here as elsewhere in Plato the appeal to Homer is ironic 
and implies that not much reliance is to be placed on what 
follows. As the illustrations proceed, we receive further indica- 
tions that Plato intends the whole performance as a reduction 
to the absurd (e.g., the references to the inspiration derived 
from Euthyphro), but Cratylus, whose theory is supposedly 
being supported and strengthened all this time, expresses no 
dissatisfaction when Socrates concludes — he says in fact 428G 
that Socrates' 'oracular utterances [are] much to his mind'. 
We may therefore suspect that the version of the natural 
Tightness theory which Cratylus would wish to uphold is not 
one which would be altogether compatible with the theory of 
Forms. This suspicion is immediately confirmed when we come 
to the passage cited at the beginning of this chapter, since we 
find there that Cratylus, while he is willing to agree with 
Socrates that names are given with a view to instruction, and 
that this instruction is an art and has its artisans, parts company 
with Socrates on the matter of the relative correctness of names, 
maintaining, as we noted earlier, that a name is either correct 
or not a name at all. Thus he is clearly not a subscriber to the 
natural Tightness theory in the version attributed to him by 
Socrates at 390D, E (which I have interpreted as a version 
involving the theory of Forms) since he denies one of its 
fundamental tenets, 'that not every one is an artisan of names'. 
On Cratylus' present view, every one who gives a name would 
be a competent name-giver, and all names would be correct — if 
the name is incorrect it is simply not a name. Cratylus' reliance 
on the fallacious argument about false speaking, therefore, 
upsets the relationship between the natural Tightness theory and 
the theory of Forms. If names are given by the expert who has 



his eye on the ideal name, then names will in fact possess a 
natural Tightness, but Cratylus, by denying the need for such 
experts (this is not explicit but is the obvious corollary of his 
denial of false names), has really also denied the necessity for the 
vision which supplies the expert with the knowledge which he 
needs to carry out his linguistic task. Even if he were to admit 
the existence of the Forms, he would have made it impossible to 
construct a theory of naming which utilizes knowledge of them, 
since to say that a name is either a correct name or not a name 
at all would be equivalent to saying that a name is identical with 
the ideal name or not a name at all. In other words, he has 
ruled out the possibility that the name is like the ideal name, or 
in some degree resembles it, and yet at the same time is other 
than the ideal name. The elimination of likeness or resemblance 
results in committing him to a situation in which he cannot 
really explain how names can convey anything at all about the 
essential nature of the things named ; he cannot do this because 
he has dispensed with any relationship between them except 
that of identity. A name which was exactly like the ideal name 
would tell us no more about the nature of the thing named than 
a second Cratylus would tell us about the nature of the original 
Cratylus (432Bff.). And if the name were in any degree 
different from the ideal name, it could not, on Cratylus' view, 
tell us anything either, since it would have ceased to exist. 28 

28 It is of interest to note that an objection to the possibility of incorrect 
names which is obviously Eleatic in origin comes from the supposed Hera- 
clitean, Cratylus. G. S. Kirk, in an article 'The Problem of Cratylus', 
American Journal of Philology, LXXII, 1951, pp. 225-53, has maintained that 
'Plato does not depict Cratylus as a convinced Heraclitean' (p. 225) and has 
noted (p. 230) the Eleatic character of the objection. Kirk's thesis has been 
attacked by D. J. Allan (in an article with the same title (A.J.P., LXXV, 
1954, pp. 271-87)) but Allan's criticism is directed almost wholly towards 
Kirk's treatment of the Aristotelian evidence concerning the relationship 
between Plato and the historical Cratylus and seems to me not to face the 
problem of what actually goes on in the dialogue. (In connection with 
Allan's paper, H. Cherniss, 'Aristotle, Metaphysics 987A32-B7', A,J.P. t 
LXXVI, 1955, pp. 184-86, should also be consulted.) I think myself that 
the undeniable combination of Heraclitean and Parmenidean elements in 
the character of the Platonic Cratylus is partly meant as one more illustration 
of the fact, which Plato sees very clearly, that the two philosophies lead 
equally to the obliteration of the distinction between truth and falsity. I 
think Plato means to say also that if one persists in trying to find out the 



In order to refute Cratylus' version of the natural rightness 
theory, then, it is clear that Plato must devise some means of 
establishing the existence of false names, and this he does 
through his analogy between naming and painting 43oBff. 

Soc. Let us see, Cratylus, if we cannot come to terms somehow. 

You would agree, would you not, that the name is one thing and 

the thing of which it is the name is another ? 
Cra. Yes, I should. 

Soc. And you agree that the name is an imitation of the thing named ? 
Cra. Most assuredly. 
Soc. And you agree that paintings also are imitations, though in a 

different way, of things ? 
Cra. Yes. 
Soc. Well then . . . can both of these imitations, the paintings and 

the names, be assigned and applied to the things which they 

imitate, or not ? 
Cra. They can. 
Soc. First, then, consider this question: Can we assign the likeness 

of the man to the man and that of the woman to the woman, and 

so forth ? 
Cra. Certainly. 
Soc. And can we conversely attribute that of the man to the woman, 

and the woman's to the man? 
Cra. That is also possible. 

Soc. And are these assignments both correct, or only the former ? 
Cra. The former. 
Soc. The assignment in short, which attributes to each that which 

belongs to it and is like it. 
Cra. That is my view. 
Soc. To put an end to contentious argument between you and me, 

since we are friends, let me state my position. I call that kind of 

assignment in the case of both imitations — paintings and names — 

correct, and in the case of names not only correct but true; 

and the other kind, which gives and applies the unlike imitation, 

I call incorrect and, in the case of names, false. 

Cratylus, while willing to concede Socrates' point about in- 
correct assignment in the case of painting, will not at once 

nature of reality by studying names, there is one sense in which it hardly 
matters what view one takes of the character of these names. The Hera- 
clitean view has, of course, some extra difficulties of its own, but the really 
important thing is to get things in the right order, i.e., to turn our attention 
first to to ovto, then to language. 



concede the point about names — he suggests that the two cases 
may be different. However, after Socrates again emphasizes 
the fact that both paintings and names are imitations, he is 
persuaded to give in 431 A. 

In this way, by getting Gratylus to admit that there may be 
names which are assigned incorrectly and are false, Socrates 
intends 'to put an end to contentious argument' between him- 
self and Cratylus. That is, he intends to get rid of the fallacious 
argument about false speaking which is doing damage to the 
natural Tightness theory in the version compatible with the 
theory of Forms. 29 

Cratylus and his theory have thus been dealt with in the 
same way as Hermogenes and the convention theory. In the 
opening pages of the dialogue (383A-384E) the two theories 
are presented as opposed to each other, and so they are in many 
ways, but there is one important thing they have in common : 
both assume that all names are true. Therefore the way to 
refute both theories is to show that, on the contrary, there are 
false names as well. 

The refutation of these two theories of language is, of course, 
one of the essential tasks accomplished in the Cratylus, but it is 
not the main purpose of the dialogue except in the indirect sense 

29 It is worth noticing that when Plato is in the process of developing the 
doctrine of recollection in the Phaedo, the principle that a painting may be 
like the object it represents is assumed to hold good 73E. If Cratylus had 
been present at the discussion, we might have expected the objection, how 
can the painting exist if it is not the thing painted ? In the Phaedo, however, 
Simmias and Cebes are both in sympathy with the theory of Forms, and this 
type of objection is not raised. The doctrine of recollection involves a 
gradual recovery of knowledge, so that at any point in the process it is 
difficult to say whether the person recollecting either knows or does not know 
a particular thing. We have already seen from the Theaetetus that the Eleatic 
insistence on a choice between knowing or not knowing is incompatible 
with the fact of memory. We can see a similar incompatibility if we compare 
the remarks on painting in the Cratylus with those in the Phaedo : the principle 
of likeness in painting is criticized in the Cratylus for reasons which stem 
ultimately from the knowing — not knowing dichotomy and the denial of 
false speaking; it is accepted in the Phaedo in a context in which there is also 
acceptance both of the fact of memory and of its constant companion, the 
theory of Forms. (It should also be noted that when the doctrine of recollec- 
tion is introduced in the Meno, it appears there as the direct answer to an 
Eleatic difficulty about the possibility of knowledge 80D. See below 
pp. 84-86.) 



in which its accomplishment lends support to the theory of 
Forms. The primary concern of the dialogue, as I see it, is to 
support the theory by answering the following three questions : 
( i ) Can the nature of reality be discovered through the study 
of language, to which the answer is quite definitely, no ; (2) 
How then can it be discovered, to which the answer is, through 
direct knowledge of ra 6Wa; and (3) What do we find out 
about t<x ovra by this method, to which the answer is, we 
find out that there are really existing things such as absolute 
beauty, etc. These three questions are treated with varying 
degrees of length and answered with varying degrees of 
definiteness, but that they comprise the substance of the dialogue 
seems to me clear. 

The arguments against the ability of the study of language to 
provide us with reliable information about the nature of reality 
are chiefly two. The first is that there seems to be no consistent 
view of reality to be gained from the study of language since it is 
possible to construct etymologies of rest which are just as 
convincing as the etymologies of motion (437Aff.). We can 
guess, of course, from Plato's remarks at 41 iB and 439C about 
the philosophers who make themselves dizzy with turning 
around and therefore suppose that reality turns too, that he 
would, if forced to make a choice, prefer the etymologies of 
rest, but that he regards the whole etymological method as 
wrong is obvious from the second and more fundamental 
objection to it which Socrates puts forward at 438Aff. : 

Soc. ... A little while ago . . . you said he who gave names must 

have known the things to which he gave them. Do you still hold 

that opinion, or not? 
Cra. I do. 
Soc. And do you say that he who gave the first names also knew the 

things which he named ? 
Cra. Yes, he knew them. 
Soc. But from what names had he learned or discovered the things, 

if the first names had not yet been given, and if we declare that it 

is impossible to learn or discover things except by learning or 

ourselves discovering the names? 
Cra. I think there is something in what you say, Socrates. 
Soc. How can we assert that they gave names or were law-givers 

with knowledge before any names whatsoever had been given, 



and before they knew any names, if things cannot be learned 

except through names? 
Cra. I think the truest theory of the matter, Socrates, is that the 

power which gave the first names to things is more than human, 

and therefore the names must necessarily be correct. 
Soc. Then, in your opinion, he who gave the names, though he 

was a spirit or a god, would have given us names which made 

him contradict himself? Or do you think there is no sense in 

what we were saying just now ? 
Cra. But, Socrates, those that make up one of the two classes are 

not really names. 

However, since Socrates has previously shown that some 
names can be derived from motion and some from rest, 
Cratylus has no way of deciding which of these two opposed sets 
are to be the true names. He is therefore constrained to agree 
when Socrates goes on to say 438E : 

... it is plain that we must look for something else, not names, 
which shall show us which of these two kinds are the true names, 
which of them, that is, show the truth of things. 

And since they have also agreed ' that names which are rightly 
given are like the things named and are images of them', he is 
also constrained to agree with Socrates' conclusion 43gAfT. : 

Then if it be really true that things can be learned either through 
names or through themselves, which would be the better and 
surer way of learning? To learn from the image whether it is itself 
a good imitation and also to learn the truth which it imitates, or 
to learn from the truth both the truth itself and whether the image 
is properly made ? 
Cra. I think it is certainly better to learn from the truth. 

In this passage Plato has finally followed out to their logical 
conclusion certain objections which appeared earlier during the 
long discussion of the various etymologies. At 409D, for instance, 
Socrates was asked by Hermogenes to explain the word 7rvp. 
' Either, ' replies Socrates, ' the muse of Euthyphro has deserted 
me or this is a very difficult word.' However, he has a con- 
trivance (fxrjxavrj) to deal with such cases; he attributes them to 
foreign sources rather than trying to explain them in terms of 
other Greek words. All that he has done, however, is to put an 
arbitrary end to the infinite series of questions which inevitably 
arises when one attempts to explain names by means of other 



names, by shifting the burden to another language. Socrates 
goes into the topic of the contrivance and its relation to an 
infinite series of explanations much more thoroughly at 42iDff. 
Here he suggests what we have probably already suspected, 
that the device of foreign origin is not the real solution of the 
difficulty, since an ancient Greek word might be identical with 
a modern foreign one. (In this case the explanations would 
necessarily continue.) Furthermore, he points out that if the 
person giving the explanations is pressed hard enough, he will 
eventually give up. The contrivance was, of course, a way of 
not giving up, but since this has failed, something else will have 
to be devised. What Socrates presently suggests is to see 
whether it might not be possible to carry the investigation back 
a stage further by making the ultimate elements letters instead 
of names. We can guess in advance that this will not work, since 
Socrates is still proceeding on the assumption (to be rejected 
438E) that the study of language in some form will finally 
succeed in showing the nature of the things named. Even if this 
method should bring us to the point at which the ultimate 
elements imitate reality, the problem is still not completely 
solved, since if language is taken as genuinely imitative, it would 
seem as if the correct way to name, say, a cock, would be by 
crowing, not by uttering the word 'cock'. As Socrates says at 

It will, I imagine, seem ridiculous that things are made manifest 
through imitation in letters and syllables ; nevertheless it cannot be 
otherwise. For there is no better theory upon which we can base the 
truth of the earliest names. 

To use the theory of foreign origin or to say that the god gave 
these names (as we have just heard Cratylus suggest at 438C) 
' are merely clever evasions on the part of those who refuse to 
offer any rational theory of the correctness of the earliest 
names' (426A). 

Socrates is of course right to insist that if the investigation of 
reality is to be conducted by means of the study of language, it 
is essential to explain correctly the linguistic elements which are 
the direct imitators of reality, since 'if anyone is, no matter 
why, ignorant of the correctness of the earliest names, he cannot 
know about that of the later'. As in a series of deductions in 



geometry, the initial error may be small, but it will permeate 
the whole series. (Socrates uses this illustration later at 436D.) 
What Socrates does not declare openly here (although he gives 
us a hint by saying 42 6B 'I think my notions about the earliest 
names are quite outrageous and ridiculous') is that the whole 
etymological procedure is unsatisfactory. We do not get his 
clear objection to it until 438A, B (quoted above), i.e., that the 
giver of the earliest names could never have given these names 
correctly if he were solely dependent on other names in order 
to learn the true nature of things, since at that time, no names 
as yet existed. 

As soon as this objection has been made, it becomes possible 
for Socrates to begin looking for 'something else, not names' 
(438D) as a criterion for judging which names are correct. 
It appears, after all, 'that things may be learned without 
names' and that the way these things are learned is 'through 
each other, if they are akin, and through themselves'. The 
names which are rightly given are like the things named and 
are images of them, but to learn whether these images are 
properly made, it will be best to learn the truth (i.e., tol ovra) 
first, rather than (as has been the attempted method throughout 
the dialogue) to attempt to learn the images first. In other words, 
the proper order has been restored to the study of language ; 
reality is to be apprehended first, after which it will be possible 
to judge whether a given name portrays this reality correctly. 

It now becomes apposite for Socrates to hazard some views 
as to the nature of this reality. So long as he continued to 
investigate the claim of language to be an index to ra 6Wa, 
what he himself thought about ra ovra was irrelevant. There 
have of course been a number of hints that Socrates prefers rest 
to motion, but he cannot embark on an exposition of this view 
until Gratylus has agreed that the correct way to learn the truth 
(and also to learn what names are properly made images of this 
truth) is from the truth itself and not from language. As soon 
as Cratylus has made this admission, however, the time has 
come for Socrates to point out the difficulties in the view that 
reality is in total flux. At 439G he speaks as follows : 

Suppose it should prove that although those who gave the names 
gave them in the belief that all things are in motion and flux . . . 
still in reality this is not the case, and the name-givers them- 



selves, having fallen into a kind of vortex, are whirled about, 
dragging us along with them. Consider, my worthy Cratylus, a 
question about which I often dream. Shall we assert that there is 
any absolute beauty or good or any other absolute existence, or 
Cra. I think there is, Socrates. 

Absolute beauty, Socrates continues, is always such as it is. 
If it were always passing away, we could never say that it was 
this or that, since it would have changed in the very instant of 
our speaking. A thing which is never in the same state cannot 
really be anything, nor, in fact, can it ever be known by anyone. 
Knowledge requires that that which knows and that which is 
known must always be — a set of conditions which bears no 
relation to the condition of flux or motion. Of course, Socrates 
concludes, the doctrine of Heraclitus and others may be true, 
but we are surely left in no doubt as to his opinion of this doc- 
trine. In any case, * no man of sense can put himself and his 
soul under the control of names, and trust in names and their 
makers to the point of affirming that he knows anything'. 
Socrates is content to have discredited the study of language 
as a means of apprehending ra 6Vra; what this reality is like, 
he insists on less forcibly, but if we have been alert throughout 
the dialogue, we can hardly fail to note that the theory of Forms 
has been advocated all along the line. 

This advocacy of the theory of Forms seems to me, as I have 
said, to be the primary purpose of the dialogue, 30 and the 
refutation of the two theories of language a subsidiary and 
auxiliary one. Nevertheless, Plato is certainly not intending to 

30 If the theory of Forms really is the dominant theme of the dialogue, it 
may appear strange that so much space has been given to etymology. It is 
clear not only from the Cratylus but from passages in other dialogues that 
Plato was interested in etymology for its own sake (cf. Protagoras 361 D; 
Phaedo 81 C; Phaedrus 237A; Theaetetus 194C; Sophist 22 iB), but he certainly 
did not expect to gain true knowledge by this means. As far as the structure 
of the dialogue is concerned, I am quite willing to agree with Meridier when 
he writes : 

On convient en general que la composition du Cratyle n'offre pas la belle ordon- 
nance ni Pequilibre si sensibles dans d'autres dialogues platoniciens. On s'est 
montre surtout choque du developpement disproportionne donne a la partie 
etymologique. Platon croyait avoir ses raisons; mais il est certain que du point de 
vue artistique Peconomie de l'oeuvre en a souffert. {Introduction to the Bude 
edition, Paris, 1931, p. 33.) 



discredit the use of language altogether. Some things, he says, 
can be learned without language, but this sort of learning is 
presumably restricted to the name-maker who has a vision of 
the Forms and thus constructs the earliest names correctly, and 
to the dialectician who is the judge of the name-maker's work. 

What Plato's own theory of language would be like, we can 
guess with, I think, a fair degree of accuracy : 

(i) It is based on knowledge of the Forms, and this knowledge 
is the source of any judgement as to whether a name is correctly 
or incorrectly made. The error of the convention theory and of 
Cratylus' version of the natural Tightness theory was that both 
theories, in addition to denying the existence of false names, put 
things in the wrong order; they attempted to discover the nature 
of reality through the analysis of names (in which case they had 
no way of telling whether the names themselves were correct) 
instead of apprehending reality first (in which case the criteria 
for correctness naturally follow). 

(2) It will insist that there are degrees of correctness in names, 
or, in other words, that there are false names as well as true. This, 
of course, is closely tied to point (1) since Plato's view that 
names, to be correct, must be given with knowledge of the 
Forms, implies, as Socrates says at 390D, that name-giving is a 
task for experts. As we have already seen, both the convention 
theory and the natural rightness theory militated against the 
need for expertness in name-giving (and therefore against any 
need for knowledge of the Forms) by arguing, each in its own 
distinctive way, that all names were true. 

(3) It will involve some theory of imitation. Socrates says at 


I myself prefer the theory that names are, so far as is possible, like 
the things named ; but really this attractive force of likeness is, as 
Hermogenes says, a poor thing, and we are compelled to employ, in 
addition, this common-place expedient, convention, to establish the 
correctness of names. 

Socrates is not meaning to say that he admires the natural 
rightness theory in the version propounded by Cratylus (since 
this version denies the existence of false names) but is indicating 
that, while the best situation would be one in which a name 



would imitate the essence of the thing named precisely, this is 
not possible. (If the name did this, it would be the thing, in the 
sense of being indistinguishable from it.) Therefore some con- 
ventional elements must be admitted — in the case of numbers, in 
fact, the names will have to be completely conventional 435B. 
Plato sees a great many difficulties in a theory of imitation. The 
Eleatic objection to images (that they are not the thing and 
therefore do not exist) is an artificial one, and this he can dispose 
of readily. But the exact way in which a name can be sufficiently 
imitative of a thing to convey its essence remains mysterious. 
We do not name a cock by crowing or a horse by galloping, 
and even if we did so, this method would not take account of 
such things as numbers. Again, naming is something like 
painting, but this analogy is not exact. The most that can be 
said is (423E) that '. . . if anyone could imitate the essential 
nature of each thing by means of letters and syllables, he would 
show what each thing really is . . .'. The precise nature of this 
imitation Plato (I think quite rightly) does not attempt to 

What Plato's theory of language might be like beyond these 
three points it is probably not possible to say. (Nor, it should be 
said, am I concerned with whether the theory is an adequate 
theory by contemporary standards.) As it is, however, the three 
points I have suggested provide us with its basic outlines: 
names are given with relation to the Forms, some are better 
than others (since some name-givers are presumably more 
skillful than others), and they in some fashion imitate the Forms. 
These points are all closely interrelated : e.g., the varying degree 
of correctness in names is related both to the degree to which 
they imitate the Forms successfully and to the necessity for skill 
in name-giving. Altogether they form a basic complex of 
doctrines into which it is possible to fit elements as diverse as 
Plato's objections to the relativism of Protagoras (385EfT.), or 
his observation that a name may contain many inappropriate 
elements and yet be a correct name 43 2 E, or his remarks 
about the effect of a small initial error on a chain of geometrical 
deductions 436D. 

Into this complex of doctrines it is clear that we may also fit 
Plato's exposition and treatment of the Eleatic fallacy against 
false speaking with which this chapter began. It was on the 



basis of this fallacy that Cratylus wished to maintain that all 
names are correct, since a name which is incorrect is simply not 
a name at all. This is of course destructive both of the require- 
ment of skill in the giving of names and of the possibility of any 
theory that names are images or imitations of Forms. The 
fallacy of false speaking, therefore, appears in this dialogue as a 
real threat both to the theory of Forms and to a theory of 
language based upon the theory of Forms. Plato has introduced 
the fallacy, with his usual dramatic skill, to illustrate once more 
the fundamental incompatibility of Eleatic monism with his 
own philosophical scheme. 



The Hippias Minor 

the passages from the Theaetetus and the Cratylus which have 
been studied in the two preceding chapters were chosen with a 
view to providing really close parallels to arguments in the 
Euthydemus since, as was stated in the introduction, this seemed 
to be a good way of establishing a few strong cases for the thesis 
of conscious use of fallacy in Plato. In the present chapter, I 
propose to diverge from this procedure slightly, and to go to a 
dialogue which, although it employs one of the fallacies pro- 
minent in the Euthydemus, equivocation, does not make use of 
exactly the same terms. (That is, the equivocations in the 
Euthydemus were primarily on 'learn' and 'is'; here they will be 
on 'good' and 'voluntary'.) It might as a result be argued that 
the use of fallacy in the Hippias Minor is not necessarily con- 
scious, since Plato might well be aware of the ambiguity of some 
terms but not of others. On the other hand, there appears to be 
independent evidence in the Hippias Minor that the equivoca- 
tions here are in fact consciously used, so that in this case there 
is no need to invoke the principle of a precise parallel with 
arguments in the Euthydemus. The Hippias Minor may therefore 
be regarded as a dialogue which extends the range of equivo- 
cations which Plato seems to have employed consciously, and 
also as strengthening the case for the more general thesis of 
conscious use. Certainly we can again see that the study of 
fallacy leads us directly to a number of the really critical philo- 
sophical problems in Plato's thought. 

The dialogue consists of a single argument (with variations) 

F 6 5 


issuing in the paradoxical conclusion that it is the good man who 
errs voluntarily. Socrates is not himself satisfied with this con- 
clusion, but, as he says in his final speech 376C, it is 'the 
inevitable result of our argument'. We may therefore find it 
profitable to examine this argument with a view to finding out 
just how Socrates has maneuvered himself (and Hippias) into a 
position which, if we know anything about his convictions, seems 
to have been completely contrary to the one he held. We may 
then inquire what Plato intended to accomplish by the produc- 
tion of the paradox. 

At the beginning of the dialogue, Hippias has just finished a 
display on the subject of Homer. Socrates, although silent at 
first, does have one or two points on which he would like to 
question Hippias; in particular, he would like to know his 
opinion of Achilles and Odysseus: which is the better man, and 
in what respect? Hippias answers 364C that of those who went 
to Troy, Homer made Achilles the bravest, Nestor the wisest, 
and Odysseus the wiliest. This answer does not meet the ques- 
tion squarely, since Hippias (a) has given Homer's opinion, not 
his own, (b) has not said, except by implication, which man 
either he or Homer considers better, and (c) has confused the 
issue by mentioning Nestor as well. Furthermore, Hippias' 
answer has not provided Socrates with quite the material he 
needs in order to bring off his initial refutation of the sophist 
at 369B. 

If we look ahead to this refutation we see that Socrates there 
points out the incompatibility in two statements made by 
Hippias: (1) that Achilles is true, and Odysseus false (and 
wily), and (2) that the true man and the false man are in fact 
the same. In order to arrive at this incompatibility it is clear 
that Socrates will need first of all to set up a strong opposition 
between Achilles and Odysseus, and, second, to contrive some 
means of destroying the opposition he has just set up. Let us see 
how he goes about it. 

Socrates first quietly drops Nestor, whom he does not need, 
and proceeds to concentrate on the contrast between Achilles 
and Odysseus. Hippias has characterized Achilles as brave and 
Odysseus as wily, but this does not oppose the two sufficiently 
for Socrates' purposes. He therefore pretends not to have under- 
stood what Hippias meant by calling Odysseus wily, and asks 



whether Homer did not make Achilles wily as well. He now 
gets the desired response, since Hippias replies 365B that 
Homer made Achilles true and simple and Odysseus wily and 
false. Socrates then proceeds to make the opposition as com- 
plete as possible by saying 365C 'Homer . . . thought that a 
true man was one man and a false man another, but not the 
same.' Since, furthermore, his purpose is to refute not Homer 
but Hippias, he next makes sure that Hippias agrees with this 
statement. Having done this, he dispenses with Homer 
altogether (on his customary grounds that it is impossible to ask 
an absent writer what he means 31 ) and addresses himself exclus- 
ively to Hippias. 

Socrates has now established one of the two statements 
necessary to produce the refutation at 369B; he has induced 
Hippias to assert that, in his opinion, Achilles is true and Odys- 
seus false, and that the true man and the false are in no wise the 

Socrates' next step is to show that while Hippias has asserted 
Achilles and Odysseus to be totally different, he can also be 
made to assert that they are completely the same. This Socrates 
does by attributing to the false man a number of properties 
which are by nature ambiguous and which therefore cut across 
the opposition between the false man and the true. The first of 
these is power : this is an ambiguous attribute since to say simply 
that a man is powerful does not specify whether he is employing 
his power for good or for evil. Socrates has masked the ambi- 
guity by the phrasing of his question; he says 365D 

Do you say that the false are, like the sick, without power to do 
anything, or that they have power to do something ? 

Hippias falls into the trap by giving the natural answer : ' I say 
that they have great power to do many things and especially to 
deceive people. ' He is clear in his own mind that by admitting 
power as an attribute of the false man, he has admitted only 
power for evil. Socrates, however, retains the attribute 
'powerful' by itself without any of its specifications. He then 
goes on to attach another ambiguous attribute to the false man, 
that of wiliness, which has been associated with Odysseus from 
the beginning. The usefulness of this move becomes immediately 

81 €f. Protagoras 347E. 



clear, since although the attribute ' wiliness ' connotes the shifti- 
ness of the false man, it also suggests the intellectual ability 
which enables such a man to accomplish his designs. Socrates 
brings this out at 365E, asking whether the false are 'wily and 
deceivers by reason of simplicity and folly, or by reason of 
shrewdness and a sort of intelligence ? ' Here again, Hippias, in 
choosing the latter alternative, probably has in mind a misused 
intelligence, but Socrates prefers to retain the attribute in its 
unspecified form, saying 365E 'they are intelligent then, as it 
seems'. After adding 'knowing' and 'wise' to the list, he 
reminds Hippias 366A that ' the true and false are different and 
complete opposites of one another' (evavTiajTarovs aAA^Aois) 
and that Hippias has now placed the false among the powerful 
and the wise. The sphere in which the false are wise and power- 
ful is that of the uttering of falsehoods, Socrates goes on; a man 
who lacks this power would not be false. At this point Socrates 
is more or less confirming Hippias in his view that to be powerful 
and wise means to be powerful and wise for evil, but we shall see 
that he soon drops this line of thought, and begins instead to 
develop these two notions in the direction of their capacity for 
good. In order not to make the transition too abrupt, he begins 
with an example which is morally indifferent: a man has 
power who is able to do what he wishes when he wishes, e.g., he 
can write another man's name if and when he wishes 366G. 
The next step is to carry the ideas of power and wisdom into 
the field of intellectual skills, e.g., arithmetical calculations. 
Hippias' admission that he is wise and powerful in these matters 
does not immediately lead to any moral judgment, but Socrates 
takes care to lay the groundwork for this addition, since he 
induces Hippias to say 366D that if he is most powerful and 
wisest in these matters, he will also be the best, (In other words, 
he will presently change the meaning of 'good', here 'good at', 
or 'skilled', to 'morally good'. See below 367C.) At the moment 
Socrates simply relies upon the latent feelings of approval 
aroused by the word 'best' to ease the progress from power to 
tell falsehoods to power to tell the truth. This is the crucial step, 
because he has now definitely crossed the boundary between 
truth and falsity; he has previously associated power only with 
the false, so that if true is really the opposite of false, the true 
should be the powerless, not the powerful. Socrates now goes on 



to make his position stronger by dwelling on the fact that the 
powers for truth and falsity in calculation are to be found in the 
same man. Furthermore, this man who 'has most power to 
speak both falsehood and truth about calculations ... is the one 
who is good in respect to them'. And no one 'becomes false in 
respect to calculation other than the good man', who is also 
powerful and true. The man who is ignorant of calculations, 
on the other hand, will be bad in the sense of lacking the skill 
to tell a falsehood when he wishes; he will in fact sometimes 
accidentally tell the truth through his badness at the subject. 32 
Socrates is now in a position to present Hippias with a 
preliminary version of the refutation which he will conclude at 
369B. (As a matter of fact the refutation is complete now, but 
Socrates refrains from calling Hippias' attention to the original 
statement about Achilles and Odysseus until he has brought 
forward further examples.) He says 367G: 

You see then, that the same man is both false and true in respect 
to these matters, and the true is in no wise better than the false? 
For he is indeed the same man, and the two are not utter opposites, 
as you thought just now. 

Hippias answers, 'Apparently not, at least not in this field.' 
Socrates then repeats the same argument for geometry and 
astronomy. No really new points emerge with the addition of 
these two new examples, but some of those implied in the pre- 
vious one come out more clearly. The bad geometer is men- 
tioned as well as the good, and it is pointed out that he will be 
powerless to speak falsehood. (Thus the opposition between 
good and bad is being more explicitly allied with that between 
true and false, and both oppositions will be destroyed together.) 
We see also that the man who lacks power in astronomy is 
characterized not only as being unable to speak falsehood but 
as being ignorant. Badness, powerlessness, and ignorance are 
all being separated off from the man who is able to speak both 

32 The notion of power is similarly ambiguous in the Hippias Major 
295Eff. : Aiivafiis pkv apa KaXov, a&vvafila 8e alaxpov; asks Socrates, thus 
identifying SiW/xis with the beautiful and the useful. But at 296G it 
turns out that power and useful things are sometimes useful for the accom- 
plishment of something bad, in which case one would hardly wish to call 
them beautiful. (The possibility that the beautiful might be powerful for 
good 296D is investigated next, but this fails on other grounds.) 



truth and falsehood, i.e., who is both false and true. Hippias 
will not, Socrates concludes, be able to find a single art, even 
of those many in which he himself is expert, in which the same 
situation does not prevail: the man who is best at truth in 
any art will also be best at falsehood. Hippias cannot give any 
exception to this rule, so he is now forced to accept the con- 
sequences: that what he has just now admitted about the 
identity of the true and the false man is directly contradictory to 
what he said earlier, in the case of Achilles and Odysseus, about 
their being different. (The inference to be drawn from the 
identity of the two is that neither is better than the other ; thus 
the original question about the relative merits of Achilles and 
Odysseus remains unanswered.) 

Hippias attempts to shy away from the refutation by com- 
plaining that Socrates is only talking about details and not 
paying attention to the whole subject. He himself wants to 
return to Homer and to prove by quotations that Achilles 
really is better than Odysseus, since one tells falsehoods and the 
other does not. He suggests that both he and Socrates should 
make speeches on this subject and should let the listeners decide 
which of them speaks better. Socrates is now willing to make at 
least a pretence of meeting Hippias on this ground, since he has 
succeeded in bringing off one refutation and can afford to 
humor his opponent briefly while he prepares the way for the 
next. He will show, by quotations of his own, that Achilles is 
just as wily and false as Odysseus. This proposal produces 
from Hippias the response which provides Socrates with the 
material for the refutation occupying the remainder of the 

Hippias, while he has to admit that Achilles does tell false- 
hoods, still considers him the better man, since what falsehoods 
he tells he tells against his will; Odysseus, on the other hand, 
tells his voluntarily and by design. If Hippias really thinks this, 
Socrates says, then he should, to be consistent, consider not 
Achilles but Odysseus the better man, since they agreed earlier 
that those who uttered falsehoods voluntarily were better than 
those who did so involuntarily. Hippias is surprised at this, 
since Socrates seems to be upsetting the common-sense views of 
forgiveness and punishment. Socrates admits that he is all 
astray about these matters — sometimes he agrees with Hippias, 



but at the moment a paroxysm of ignorance has seized him, and 
those who do evil voluntarily seem to be better than those who 
do it involuntarily. 'And,' he says, 'I lay the blame for my 
present condition on the previous argument' (372E). 

This is the first really open hint in the dialogue that the 
conclusion which has been reached at the end of the first 
refutation is no more acceptable to Socrates than it was to 
Hippias. Socrates is, however, perfectly clear about the cause 
of this conclusion; it is the argument which is to blame. Since 
he has constructed the argument himself, he has certainly not 
placed himself in this predicament by inadvertence. It seems 
then as if it must be at least part of Plato's purpose to call 
attention to the way in which the argument has been con- 
structed. We shall see further reason to think this as the dialogue 

Next is a brief interlude during which Hippias tries to back 
out of the discussion on the ground that Socrates always makes 
confusion and trouble in arguments (373Bff.) Socrates, however, 
is still eager to investigate the matter they have just mentioned 
(which are better, those who err voluntarily or those who do so 
involuntarily), and Hippias is finally persuaded to go on 
answering questions. 

Socrates no longer continues to remind Hippias that the 
proposition they are now considering is the one that resulted 
from the first argument, but we shall see that the new argument 
repeats one part of the first one all over again. The terminology 
is different ('voluntary' takes the place of 'powerful') and the 
illustrations are broader and are presented in a way which 
brings out their moral connotations, but the examples leading 
to the conclusion that the good man errs voluntarily are con- 
structed on the same lines as the example of, for instance, the 
calculator who has the power to make mistakes. 

The shift in the character of the examples may be seen at once. 
Socrates has previously restricted himself to intellectual error. 
Now he begins to speak of bodily actions, a sphere in which it is 
more natural to think of mistakes as a disgrace. The man who 
runs slowly, for instance, does a disgraceful act. If he does this 
voluntarily, however, he is better (and again, of course, 
'better' means more skillful) than the man who does it involun- 
tarily, that is, than the man who runs slowly because he lacks 



the power to run any faster. Hippias, as in the earlier discussion, 
admits that Socrates is right in this particular case, but he does 
not immediately concede the general principle. It is thus neces- 
sary for Socrates to continue the illustrations at some length, 
moving from bodily actions to the senses and from there to 
possessions of various sorts, some attached to the body (such as 
the feet) and some separate from it (such as various kinds of 
instruments, both animate and inanimate). The fact that this 
argument is the same as the first shows clearly when the 
illustrations swing round again to the intellect. We have (e.g., at 
375B) the example of the archer: it is better to possess the mind 
which voluntarily misses the mark than the one which does so 
involuntarily. Socrates repeats the point in the case of medicine 
and of music ; in these fields it is again better to possess a mind 
which errs voluntarily rather than the reverse. Hippias, 
however, still rebels against the universal conclusion; he says 
at 375D: 

But it would be a terrible thing, Socrates, if those who do wrong 
voluntarily are to be better than those who do so involuntarily. 

Socrates' answer is agairt to remind him of the source of this 
view, saying, 'But surely they appear to be so, from what has 
been said 5 (e/c ra>v elprjfxevojv). 'Not to me', replies Hippias. 

Socrates then gives one further example, the crucial instance 
of justice 375D, E. This, he suggests, is either a kind of know- 
ledge or a kind of power or both. If it is a power, then the more 
powerful soul is the more just, since they have already agreed 
that the powerful soul is better than the one which lacks power. 
And if it is a kind of knowledge, the wiser soul is more just than 
the one which is ignorant. This more powerful and wiser soul is 
not only better but more able to do both good and disgraceful 
acts in every sphere. 33 When it does disgraceful acts, it does 
them voluntarily by reason of power and art, either or both of 
which are attributes of justice. The doing of disgraceful or evil 
acts is the same as doing injustice, so that when the powerful and 
better soul does these acts it will be doing injustice voluntarily. 

33 Cf. the idea put forward in Republic I 334A that the just man is the 
best thief. (Here there is ambiguity in the adjective Bcivds; but see also the 
analysis of D.J. Allan in the notes to his edition of Republic I, London, 1953, 
pp. 89-90.) 



This powerful and better soul is the good soul and is the pro- 
perty of the good man. Socrates thus concludes that 

he who voluntarily errs and does disgraceful and unjust acts, 
Hippias, if there be such a man (etirep tLs eoriv ovtos), would be 
no other than the good man. (376B.) 

Hippias is still unable to agree with Socrates, nor, in fact, is 
Socrates able to agree with himself; again this situation is the 
'inevitable result of [the] argument' (376G). Socrates wanders 
up and down on this matter and is still doing so as the dialogue 

If we take Socrates' hint and examine the argument, we can 
detect the chief sources of error as before. In the first place, the 
strong type of opposition which occurred earlier between the 
true man and the false is here continued, but is replaced by a 
comparable opposition between the good man and the bad. 
(As a result the ethical connotations are wider.) It continues 
also to be the case that other oppositions follow, as, for instance, 
the all-important one between the voluntary and the involun- 
tary. (The voluntary takes the place held by the powerful in 
the earlier sequence.) When particular illustrations are given, 
other oppositions are used which are appropriate to the matter 
in hand, e.g., running quickly vs. running slowly, and so forth. 
In fact the whole notion of conducting the argument in terms 
of exhaustive alternatives is simply taken for granted, and 
Socrates makes no further attempt to call our attention to the 
procedure. (Earlier, at 366A, he took pains to point out that 
the true and the false were different and in no way the same.) 

Now these oppositions are not themselves arrived at by 
fallacious means, nor is it in itself a fallacious procedure to 
employ them in an argument. (In the Theaetetus passage, the 
knowing — not knowing dichotomy was set up by the use of 
secundum quid, but here this is not the case.) Hippias claimed 
almost from the beginning that Achilles was true and Odysseus 
false, and appeared to mean, with a little prompting from 
Socrates, that the two were completely opposed in this. 
Furthermore, if such terms are employed in a consistently 
absolute sense throughout the argument, there is no more 
deception in the conclusion than there was in the premises. Or, 
to put it another way, if the opposition between true and false 



or between good and bad is felt to be a misrepresentation of 
reality (on the ground, say, that no absolutely true or false or 
good or bad man exists), then the misrepresentation permeates 
the entire sequence; it is not something which emerges to 
confront the answerer unexpectedly in the conclusion. Therefore 
any objection to the opposition between e.g. the true and the 
false should be made straightway as soon as this opposition 
appears in the premises. On the other hand, and this is the 
important point, it may also be said that these oppositions are 
essential to the success of the ultimate refutations of Hippias 
(and in fact to any refutation) and so might be said to be one of 
the sources of error. This point is made clear by Aristotle, who 
says (On Sophistical Refutations i68a37) ' refutation is a proof of 
the contradictory 5 (o yap eAey^os* avXXoyiafAos avrufiaaeoos) . 
In other words, if Socrates had attempted to work with the 
statements 'Achilles is somewhat true', or 'Odysseus is more or 
less false 5 , he would not have been able to bring off a complete 
refutation, since refutation is effected by asserting the exact 
opposite of a statement, not its partial opposite. Thus there is, 
one might say, a certain static quality necessarily to be found 
in formal argument, and this static quality is by its very nature 
incompatible with the correct representation of things in flux. 
If however the necessity and extent of this misrepresentation be 
understood, no deception need result. Formal logic can never 
cope with becoming in all its phases, and this fact simply has 
to be accepted and faced. 

If we move on to consider the argument further, we shall see 
that the word 'good 5 is clearly used equivocally. Socrates is 
working towards the establishment of the paradox that it is the 
good man who errs voluntarily. The paradox is a paradox 
because voluntary error conflicts with virtue in the public mind. 
(Hippias has pointed out that the laws are constructed on the 
opposite assumption; it is the voluntary evil-doer who is 
punished, and he is presumably regarded as evil, not good, in 
the eyes of the law.) 'Good 5 is certainly intended to mean 
' morally good 5 in this conclusion. When the illustrations begin 
at 373G, however, with the example of the runner, it is equally 
certain that 'good 5 means 'good at 5 or 'skilled 5 . The way is 
paved for the shift in meaning almost immediately, when Soc- 
rates says at 373E, 'Then he who runs badly performs a bad 



and disgraceful (alaxpov) act in a race'. He continues to 
associate ' disgraceful ' with ' bad ' at intervals as the illustrations 
proceed, saying for instance at 375G that in connection with all 
the arts and sciences the mind is better ' which voluntarily does 
bad and disgraceful things and commits errors'. By the time the 
example of justice is reached at 375D, Hippias is well prepared 
to agree with Socrates when he associates the good man with 
the performance of disgraceful acts, although he is still unwilling, 
in spite of the argument, to allow that such acts are done volun- 
tarily. But the good man is not, of course, strictly analogous to 
the good runner or the good musician. The good runner can be 
skillful at running without necessarily being morally virtuous, 
but the good man must be virtuous since no other activity than 
the activity of being a man has been specified for him to be 
skillful at. In other words, in the absence of any other action, he 
can only be described as good because he is a good human 
being, and this sort of goodness implies virtue rather than skill. 
The term 'voluntary' is also equivocal. 34 This is not altogether 
obvious when the term is associated with good acts, but becomes 
so when it is also associated with disgraceful acts or mistakes. 
The paradoxical conclusion to the present argument is 'the 
good man errs voluntarily', a statement which is similar to 'the 
good calculator has the power to make mistakes' in the first 
part of the dialogue. The voluntary means primarily 'what is in 

34 Here I have followed Taylor's lead (Plato: the Man and his Work, p. 37) 
and have consulted Proclus. The relevant passage is 

ttXt)v oti K.V.T avrov ovk eort to €Kovatov /cat to i<f>* rjplv tccvtov. aXXa to pkv e/covatov 
iv fxovois iorlv Tots dyadois, €iTrep clkovoios /cat d^ovXrjTos 6 tojv /ca/caiv fiios' i(f>' f}plv 8e 
/cat cltto tojv rjfi€T€pcov alpeoetov /cat tcc dpapTrfpaTa. /cat yap ravra alpovp.e8a /cat 
iXopevoi Trp6.TTop.ev, dXXd Si' ayvotav. et ovv tcc apapT-qpara a/coucrta p4v, iv Tat? 
alpioeatv Se r)pwv ioTt. /cat Tavra, 7ravTa pkv tcc e'/coyata /cat aipcTa BiQTrovOev €otiv, ov 
TrdvTd Se t<x atpeTa e/couota. 

(In Platonis Rem Publicam Commentarii, ed. Kroll, Leipzig 1899, 2 vols., vol. 2, 
P- 355-) 

'However he says that the voluntary and the "what is in our power" are 
not the same. For the voluntary is concerned with goods only (since a life 
of evils is involuntary and unwished for) but mistakes are also in our power 
and arise from our own choices. We indeed choose these, and, having chosen 
them, do them, but through ignorance. If, then, mistakes are involuntary, 
and even these are among our choices, all voluntary things, I suppose, are 
also chosen, but not all things chosen are voluntary." (My translation.) 



our power', but it has also the connotation of what we wish 
or desire. Now in the example about the calculator there was 
no suggestion that qua calculator he could possibly wish to 
make a mistake; a situation in which he might be called upon 
to do so would be highly artificial and hypothetical only. 
Again in the case of the wrestler : presumably he has the power 
to allow himself to be thrown, but qua wrestler he cannot really 
desire to be thrown because he would lose the match. If we 
carry this notion over into the case of the good man, we can 
see that the situation is again hypothetical: the good man has 
the power to err, but cannot, by the very description of his 
nature, have the desire to do so. (How can a good man desire 
what is bad any more than can a fast runner desire to run 
slowly?) Plato gives us the key to all this when he says in 
Socrates' next to last speech, 'if there be any such man'; we 
are to infer that there is none. 35 If the good man performs only 
good acts, it is correct to describe him as performing them 
voluntarily since there is no conflict between what is in his 
power and what he desires. (The former classification would 
simply be broader.) But if (presumably through ignorance) he 
performs bad acts, then his power and his desire do conflict, so 
that the term 'voluntary' becomes no longer adequate to 
describe the situation. Thus it is not used in the same sense in 
both cases, and may be said to be an equivocal term. 

The means by which Socrates has produced the final paradox 
of the dialogue are thus as follows : ( i ) the oppositions between 
good and bad, voluntary and involuntary, etc., (2) the equivocal 
use of 'good', (3) the equivocal use of 'voluntary'. Of these, (1) 
is not strictly a fallacious procedure, although it is an essential 
factor in the argument, since refutation is based on contradic- 
tion. (Hippias has wished to maintain that the bad are those 
who err voluntarily and the good are those who err involun- 

35 Cf. A. E. Taylor, Plato: the Man and his Work, p. 37: 

' On reflection we see that the key to Plato's meaning is really supplied by one 
clause in the proposition which emerges as the conclusion of the matter: "the man 
who does wrong on purpose, if there is such a person, is the good man." The insinuation 
plainly is that there really is no such person as ' ' the man who does wrong on purpose, ' ' 
and that the paradox does not arise simply because there is no such person. ' (his 
italics). Cf. also Paul Shorey, What Plato Said, Chicago, 1958, p. 89, to the same 
effect. Shorey notes further, p. 47 1 , that Plato employs a similar device at Euthyphro 
7D ('If the gods quarrel') and at Gorgias 480E ('If one ought to harm anyone'). 



tarily; he is finally forced into a position in which he has 
assented to statements which imply the exact opposite.) 
(2) and (3) are technically fallacious, however, so that there is 
no doubt that the paradox has been produced by sophistical 

What has occurred in the dialogue is this: Socrates has 
refuted Hippias by unfair means. Socrates knows that the 
refutation has been unfair and understands in what its unfair- 
ness consists. Furthermore, the paradox which completes the 
refutation is contrary to his own opinion, and he understands 
and admits this. We are now in a position to ask what Plato 
intends to accomplish by means of this remarkably indirect 

In order to begin to answer this question it should first be 
noted that the two terms which are used equivocally are both 
terms which are central to Socratic ethics. It may even be 
said, further, that the distinctive character of his ethics is 
expressed by means of these particular equivocations. 

The two meanings of the word ' good ' are, as we have seen, 
'good at', or 'skilled', and 'morally good'. The good runner is 
the skillful runner, whereas the good man is the virtuous man. 
But, if virtue is to be knowledge (and there is certainly reason 
to believe that this is the Socratic view), there is a sense in which 
the two meanings will always be combined. That is, there will 
be a moral quality associated with good running, and a degree 
of technical skill associated with the performance of good deeds. 
Then again, the term ' voluntary ' was equivocal in that it was 
used to mean both ' that which is in a man's power and which 
he also desires', and 'that which is in a man's power but which 
he does not desire'. (In the latter case he is deceived, through 
ignorance of his real interests.) This equivocation is also tied 
closely to one of the major Socratic doctrines : the view that no 
man errs voluntarily. Behind this doctrine are two assumptions : 
(1) that a man can do only what is in his power, and (2) of the 
things which are in his power he will in fact do only those which 
he desires, i.e., which seem good to him. The only possible 
explanation of error, then, is that it is done through a desire for 
what is not in fact good but which merely seems good. In other 
words, error is unknowing or involuntary. Therefore, whenever 
Socrates says, in his various illustrations, such things as that the 



good wrestler is thrown voluntarily, or that good ears are 
voluntarily deaf, he is, in terms of his own doctrine, uttering 
just as much of a paradox as he is when he states in the con- 
clusion that it is the good man who errs voluntarily. To say 
that any error is voluntarily chosen shows immediately that the 
word 'voluntary' lacks its complete meaning in the given 
context; the power is present but the desire cannot be, except 
through the fact of an accompanying ignorance. The equivoca- 
tion therefore serves to show that the whole question of error is 
tied to the question of knowledge, which in turn, of course, is 
tied to the question of virtue. 

The chief point of the Hippias Minor therefore seems to me to 
be to call attention to the distinctive doctrines in the Socratic 
ethics. Plato has, as usual, adopted a highly individual method 
of effecting his purpose. He has employed arguments which, 
if we follow them out carefully, are clearly fallacious, yet he has 
used these very arguments to force us to consider the possibility 
that the equivocal character of certain terms may cast some 
light on ethical questions of the most fundamental sort. There 
are of course other things accomplished in the dialogue : we 
see, for instance, that Hippias is hopeless as a dialectician, that 
the use of exhaustive alternatives is essential to the refutation of 
a statement, and that equivocation is a fallacy which Plato 
handles with ease. But Plato's real accomplishment in the 
dialogue is, by means of a paradox which is a shock to our 
common sense, to compel us to consider what the terms ' good ' 
and 'voluntary' really mean. This is a use of fallacy which 
serves to construct, not, as in the Euthydemus, to destroy. 36 
If the constructive nature of the arguments is not as obvious as 
in the Theaetetus, this is because Hippias, as a dramatic character, 
is not capable of making the distinctions which would bring 
this out. Moreover, Socrates presumably has little interest in 

36 Aristotle writes that the solutions of fallacious arguments are 

useful for philosophy for two reasons. In the first place, as they generally turn 
on language, they put us in a better position to appreciate the various meanings 
which a term can have and what similarities and differences attach to things and 
their names. Secondly they are useful for the questions which arise in one's own 
mind; for he who is easily led astray by another person into false reasoning and 
does not notice his error, might also often fall into this error in his own mind. 

(On Sophistical Refutations, i75a5ff., trans. Forster.) I should guess that 
Plato would have agreed with both points. 



the education of Hippias, so that Plato does not represent him 
as making the distinctions either. But the necessary hints are 
there, and it is the reader's responsibility to heed them. 37 

37 The authenticity of the Hippias Minor is attested to by Aristotle at 
Metaphysics A I025a6ff. in the course of his discussion of the meanings of 
* false ' (if/€vSos) : 

Hence the proof in the Hippias that the same man is false and true is misleading; 
for its assumes (a) that the false man is he who is able to deceive, i.e. the man who 
knows and is intelligent; (b) that the man who is willingly bad is better. This 
false assumption is due to the induction ; (S«x ttjs irraycoyrjs) for when he says that 
the man who limps willingly is better than he who does so unwillingly, he means 
by limping pretending to limp. For if he is willingly lame, he is presumably worse 
in this case just as he is in the case of moral character (trans. Tredennick) . 

Aristotle's comment is worth considering on its own account. The passage 
taken as a whole illustrates the fact that Aristotle conceives the dialogue as 
consisting, as I have said, of a single argument, since he cites as contributing 
to the first refutation (that the same man is false and true) not only (a), 
which is drawn from the first discussion (especially 366A), but also (b), 
which is drawn from the second (especially 374B). In other words, he takes 
the conclusion that the same man is false and true as substantially equivalent 
to the conclusion that the good man errs voluntarily. (This could be 
rephrased as 'the same man is both good and bad'.) Further in (a), Aristotle 
notes the connection between power and knowledge and relates this to the 
first refutation. (We saw that the notion of power, when first introduced, was 
understood by Hippias to mean power for evil; the introduction of shrewd- 
ness 365E prepared the way for a shift to a power connected with good, i.e., 
with knowledge. It was by means of this shift that the false man was iden- 
tified with the true.) In (b), Aristotle has an interesting comment to make 
on the argument which Socrates has used to reach the conclusion that ' the 
man who is willingly bad is better'. This false assumption, he says, is due to 
the induction ; by this I take him to mean that the example about voluntary 
bodily error should not have been used to lead to a conclusion involving 
voluntary moral error since the two types of activity are not strictly com- 
parable. (That is, in the example about limping a pretense is made, which 
seems not to be the case in the conclusion.) If the man were in fact willingly 
lame (I think Aristotle means through having deliberately inflicted an 
injury upon himself) the induction would hold in the sense that bodily and 
moral error could be compared, but the conclusion would be that the man 
committing moral error would be worse, not better. Aristotle, it must be 
remembered, rejects the Socratic paradox that no man errs voluntarily (see 
e.g., Nichomachean Ethics 1 136a! ff.) ; thus we have a situation in which he and 
Plato, although holding opposite views on the question of voluntary error, 
come both to reject the conclusion (b) (that the man who is willingly bad is 
better) : Aristotle because he objects to 'better', Plato because he objects to 
'willingly bad'. 



in the preceding chapters I have been chiefly concerned with 
some of the ways in which Plato has made use of fallacy in 
connection with the exposition of his own philosophical views. 
But I have also asserted (i) that Plato possesses a complete 
mastery of certain kinds of fallacy (especially the fallacies of 
equivocation and secundum quid), and (2) that on occasion he 
deliberately employs such fallacies as an indirect means of 
supporting his own position. 

Of these two points, the first is more immediately palatable 
than the second. It is one thing to credit Plato with an under- 
standing of certain points of logic, but quite another to assert 
that this understanding is put to what might be regarded as 
questionable use. If fallacies are, after all, bad arguments, it 
might be asked whether it is fair play on the part of Plato to 
employ them, even for the most laudable ends. 

There is a real dilemma here. In the first place, there is 
certainly no doubt that fallacies do occur in the dialogues. If so 
(to argue for a moment in Eleatic style) , Plato either knew this 
or he did not. If he knew it, he lays himself (and Socrates) open 
to the charge of immoral practise in argument; if he did not, he 
is apparently ignorant of some fairly elementary parts of logic. 38 

38 Sometimes, of course, it is argued that Plato could not have been 
expected to know much logic at this stage of the world's history. See, for 
instance, E. R. Dodds in his commentary on Gorgias 474C4-476A2 
(Oxford, 1959, p. 249) who takes this line, but not without a certain 
uneasiness : 

But while Polus' view is muddled and ultimately untenable, Socrates' formal 
' refutation ' of it seems to turn merely on the ambiguity of the word ci^eAi/xov. 
When Polus said that doing wrong was less admirable, he clearly meant that it was 



In this book I have argued that the evidence of the Euthy- 
demus is too strongly on the side of Plato's conscious use of fallacy 
for his understanding of the fallacies in that dialogue to be 
denied. I have then proceeded to argue that when similar 
fallacies occur in subsequent dialogues, they continue to be 
understood and their use is still deliberate. On the other hand, 
I have also argued (particularly at the close of the Theaetetus 
chapter, pp. 43 ff.) that arguments which may, from their 
resemblance in form to arguments in the Euthydemus, be called 
'eristic', need by no means be employed in an eristic spirit. 
In this way I have tried to uphold Plato's intelligence without 
representing him as a logical twister, but I am actually less 
concerned to defend his character than to describe what he 
appears to be doing in certain dialogues. Whether what he does 
is reprehensible can only be decided by the individual reader, on 
inspection of the passages involved. The whole issue would in 
fact be trivial if it were not that I suspect strongly that it is 
precisely this reluctance to pass an adverse moral judgment 
upon Plato (or perhaps more especially upon Socrates) which 
has prevented scholars from considering seriously the possibility 
that Plato's use of fallacy really is deliberate. The result has 
been that Plato's competence as a logician has failed to be 
evaluated correctly — a matter which is far from trivial. 39 

less <b^4Mfiov for the community and from this it does not immediately follow that it is 
less d)<f>4\i[jLov for the agent, i.e. kolkiov in Polus' sense of that term. The underlying 
thought is, no doubt, that since regard for justice is recognized even by Polus as 
KaXov, as evoking immediate admiration, it must be a necessary constituent of 
the 'good' or happy life. But Plato has obscured this point for the sake of giving 
his argument the appearance of a formal proof. We must remember that when the 
Gorgias was written the study of logic was still in its earliest infancy (as Aristotle's 
Sophistici Elenchi sufficiently shows). Nevertheless, it is not easy to believe with 
T. Gomperz and others that Plato was wholly unconscious of the equivocation .... 

39 Of course it is doubtful whether the Greeks themselves would have been 
much troubled by the moral issue. E.g., Aristotle in On Sophistical Refutations 
(a treatise in which, incidentally, he gives some excellent advice on how to 
be a sophist of the eristic type, i72bio-i73a3i and 174a 17- 174.D40) writes 
as follows : 

In the first place, then, just as we say that we ought sometimes deliberately to 
argue plausibly rather than truthfully, so too we ought sometimes to solve questions 
plausibly rather than according to truth. For, generally speaking, when we have 
to fight against contentious arguers, we ought to regard them not as trying to 
refute us but as merely appearing to do so; for we deny that they are arguing a 
case, so that they must be corrected so as not to appear to be doing so. For if 
G 8l 


As I have said in the introduction, the question of Plato's 
consciousness of the fallacious character of a given argument 
needs to be settled before proceeding to a discussion of the 
way in which that fallacy is used. But it may be of interest, 
at the conclusion of this study, to examine brief passages 
in additional dialogues in which it seems highly probable, if not 
completely settled, that Plato knew quite well what he was 

I. Laches igoBff. 

The dialogue has begun with the question whether the art of 
fighting in armor should be learned by young men, but has now 
come round to inquire about the nature of virtue 190B. It will 
be easier, Socrates says, to investigate a part of virtue rather 
than the whole — why not consider the particular virtue which 
the art of fighting in armor is supposed to produce, i.e., courage ? 
Laches is asked to tell what courage is, but, like many of those 
who converse with Socrates, he does not at first understand the 
nature of a general definition. After one false start, however, he 
suggests that courage is 'a sort of endurance of the soul' 
(Kaprepla ns rrjs ijjvxrjs 192B). Socrates argues against the 
proposed definition in the following way: courage presumably 
is a noble thing. With respect to endurance, however, wise 
endurance is good and noble, but foolish endurance evil and 
hurtful. It seems then, as though only wise endurance should be 
courage. Yet there appear to be cases in which we regard a man 
who endures foolishly (e.g., a soldier in the face of overwhelming 
odds) as braver than the man who endures with sense. To call 

refutation is unequivocal contradiction based on certain premisses, there can be no 
necessity to make distinctions against ambiguity and equivocation; for they do 
not make up the proof. But the only other reason for making further distinctions is 
because the conclusion looks like a refutation. One must, therefore, beware not 
of being refuted but of appearing to be so . . . (i75a32ff., trans. Forster). 

Furthermore it seems clear that Aristotle regarded the respondent in a 
dialectical argument as entirely responsible for his own defense : ' since the 
right to draw distinctions is conceded in arguments, it is obvious that to 
grant the question simply, without making distinctions, is a mistake' 
1 75b3off. The opponents of Socrates, when they are refuted, quite regularly 
are so because of this very failure to draw the right distinctions ; we need not 
fail to draw them, however, since Plato has on most occasions provided the 
means for us to do so. 



this man 'courageous 5 would contradict the original statement 
that courage is noble but foolish endurance is ignoble. 

Laches admits the contradiction but can see no way of 
escape. There is in fact no escape so long as he continues to call 
'courageous' actions which should more properly be called 
'rash'. After Nicias enters the discussion, however, this distinc- 
tion is ultimately made. 

Nicias offers ( 1 95 A, 1 96C) a definition of courage as 'know- 
ledge of the grounds of hope and fear' (rj tow Sclvwv koX 
OappaXeojv imarqfirj) . Among other arguments against this 
definition, Socrates puts forward one which is basically the same 
as part of his objection to the definition offered by Laches: 

... I conceive it is necessary for him who states this theory to 
refuse courage to any wild beast, or else to admit that a beast like 
a lion or a leopard or even a boar is so wise as to know what only 
a few men know because it is so hard to perceive. Why, he who 
subscribes to your account of courage must needs agree that a 
lion, a stag, a bull, and a monkey have all an equal share of 
courage in their nature. 
Lack. Heavens, Socrates, how admirably you argue ! Now answer us 
sincerely, Nicias, and say whether those animals, which we all 
admit to be courageous, are wiser than we are; or whether you 
dare, in contradiction of everyone else, describe them as not even 
courageous. (196E.) 

Nicias, in response, makes the answer which Laches should 
have made in the case of the foolhardy soldier, etc., in the 
earlier passage; 

No, Laches, I do not describe animals, or anything else that from 
thoughtlessness has no fear of the dreadful, as courageous, but 
rather as fearless and foolish. Or do you suppose I describe all 
children as courageous, that have no fear because they are thought- 
less ? I rather hold that the fearless and the courageous are not the 
same thing. ... So you see, the acts which you and most people 
call courageous, I call rash, and it is the prudent acts which I 
speak of that are courageous. (i97AfF.) 

Socrates has succeeded in pushing the popular (equivocal) 
use of ' courage ' so far that he has at last forced his opponent to 
separate off one meaning of the term, i.e. rashness. (In the end, 
courage becomes a divine type of knowledge and is thus not 
simply a part but the whole of virtue 199E.) The way in which 



he has achieved the separation is by ignoring it himself and by 
proceeding as if the popular use of courage were the correct one. 
Laches can see that something is wrong, he is not sure what; 
Nicias sees the trouble and states it clearly. Thus deliberate 
ambiguity on the part of Socrates results in the clarification of 
this same ambiguity. The dialogue, like others of the same 
period, finishes with an admission of defeat 199E, but it is 
certainly not the case that nothing has been accomplished, 
since we have at least been made to see that if courage is a virtue 
it cannot be foolish. This may result in the absorption of courage 
into virtue as a whole (the problem of the Protagoras) but pro- 
gress has been made in the direction of a distinction between 
popular courage and the courage proper to the philosopher. 
Plato has made use of the fallacy of equivocation in a construc- 
tive way. 40 

II Meno 8oDfT. 

Even after some discussion of the nature of virtue, Socrates 
still does not know what virtue is. He asks Meno to join him in 
the inquiry, but Meno has an objection to the project: 

Why, on what lines will you look, Socrates, for a thing of whose 
nature you know nothing at all ? Pray, what sort of thing, amongst 
those that you know not, will you treat us to as the object of your 
search? Or even supposing, at the best, that you hit upon it, how 
will you know it is the thing you did not know ? 

Socrates is quite familiar with this sort of thing : 

I understand the point you would make, Meno. Do you see what a 
captious argument you are introducing — that, forsooth, a man 
cannot inquire either about what he knows or about what he does 
not know ? For he cannot inquire about what he knows, because he 

40 Plato fairly frequently employs the method we have just seen Socrates 
using here in the Laches, that of taking a term in such a wide sense initially 
that the opponent will finally insist on having the extension limited. We 
may compare Gorgias 489 B, G where Socrates forces Callicles to say that by 
' superior ' (Kpdrrcov) he does not mean ' physically stronger ', and Republic I 
338C, D where a similar treatment is administered to Thrasymachus. (In 
these cases, however, the limitation of meaning results in clarification of 
the opponent's wrong view, not that of Socrates.) In his note on the latter 
passage Shorey writes, ' To the misunderstanding of such dramatic passages 
is due the impression of hasty readers that Plato is a sophist' (Loeb edition, 



knows it, and in that case is in no need of inquiry; nor again can he 

inquire about what he does not know, since he does not know about 

what he is to inquire. 

Men. Now does it seem to you to be a good argument, Socrates ? 

Soc. It does not. 

Men. Can you explain how not ? 

Soc. I can (8oDff.) 

Socrates then goes on to outline the doctrine of recollection 
(av&iiviqcHs) , throwing special emphasis upon the fact that 
this doctrine is one which makes learning possible : 

For as all nature is akin, and the soul has learned all things, there is 
no reason why we should not, by remembering but one single thing 
— an act which men call learning — discover everything else, if we 
have courage and faint not in the search; since, it would seem, 
research and learning are wholly recollection. So we must not 
hearken to that captious argument: it would make us idle, and is 
pleasing only to the indolent ear, whereas the other makes us 
energetic and inquiring. Putting my trust in its truth, I am ready to 
inquire with you into the nature of virtue. (8iCff.) 

The questioning of the slave boy then follows by way of illus- 
tration. The investigation of the nature of virtue, however, does 
not proceed until Socrates has again spoken out against the 
evil effects of Meno's argument in the strongest terms : 

. . . Most of the points I have made in support of my argument are 
not such as I can confidently assert; but that the belief in the duty 
of inquiring after what we do not know will make us better and 
braver and less helpless than the notion that there is not even a 
possibility of discovering what we do not know, nor any duty of 
inquiring after it — this is a point for which I am determined to do 
battle, so far as I am able, both in word and deed. (86Bff.) 

Like arguments in both the Euthydemus and the Theaetetus, 
this argument of Meno's at 8oD is based upon an Eleatic dis- 
tinction between knowing and not knowing which is ultimately 
destructive of the possibility of acquiring any new knowledge. 
The doctrine of recollection is introduced as a direct answer to 
an eristic trick, a trick which would stultify the desire to learn 
through the assertion that learning is impossible. The doctrine 
of recollection is, of course, as may be seen more clearly in the 
Phaedo, closely connected with the theory of Forms, so that the 
fact that Plato chooses to meet the eristic argument in the way 



he does is further evidence of the constant clash between the 
Eleatic metaphysics and the Platonic. Fallacy has been used by 
Plato, as it so often is, to give emphasis to some important 
positive doctrine of his own. 41 

The way in which the fallacy is introduced, however, differs 
from the method of the Laches in two important ways : ( i ) It 
occurs not in the mouth of Socrates but of one of his interlocutors. 
(In this respect the passage may be compared with Cratylus 
429B1T. where it is Cratylus, not Socrates, who puts forward 
the arguments against false speaking.) (2) It is immediately 
recognized as a fallacy and tagged as such. The result is that 
the difficulties about consciousness and culpability really do 
not arise. 

The fact seems to be that the acceptance of the thesis that 
Plato's use of fallacy is deliberate presents difficulties chiefly 
when fallacious argument occurs in the mouth of Socrates 
(or of some other speaker whom we tend also to regard as spokes- 
man for Plato, e.g. the Eleatic Stranger). This situation is a 
natural by-product of the dialogue form. One's instinct when 
reading the account of an argument is to take sides, and to 
assume, further, that the author has taken sides also. In the 
case of a Socratic dialogue, it seems obvious that the side taken 
by the author will be that of Socrates. If, then, we find Socrates 
(in spite of his supposedly good character) tricking an opponent 
by fallacious means, we may be inclined to shift our sympathies 
to the other side, to the side of the person being tricked. (Or, as 
I have already suggested, if we prefer to think that Socrates 
would not consciously behave in such a way, we may decide 
instead that he, and probably Plato also, were ignorant of the 
logical issues involved.) But all this taking of sides, although a 
tribute to Plato's dramatic skill, is really beside the point. The 
way to read a Socratic dialogue is not only to ask one's self, what 

41 The philosophical importance of this passage in the Meno is well 
brought out by Bernard Phillips in an article ' The Significance of Meno's 
Paradox' {Classical Weekly, vol. 42, no. 6, December 20, 1948, pp. 87-91). 
He writes, for instance, 

The paradox is evidently of sophistic origin, and though in the hands of lesser 
sophists and laymen it served only as a piece of verbal trickery, the argument itself 
expresses a philosophical point of view which is opposed fundamentally to the 
whole conception of the nature of knowledge and the function of reason which 
Plato is advancing. 



would I have said if I had been in the place of Hippias or 
Cratylus or Protagoras, but also (and this is more important), 
what has Plato actually meant to say to me, the reader, by 
means of the entire conversation ? If we read the dialogues in 
this way, there is much less likelihood of confusing ourselves by 
extraneous moral judgments, for instance: 'Plato never gives 
Hippias anything intelligent to say, so naturally Socrates wins 
the argument ; all this is very unfair. ' Of course if Plato had 
written prose treatises, none of these problems would arise, but 
he did not write prose treatises for the very good reason that he 
conceived the exercise of philosophy to be the exercise of 
dialectic — hence the only way to give even a semblance of 
philosophical activity in written form was to write dialogues. 
(That he did write dialogues is the Platonic scholar's joy and 
also his despair.) 

The whole question of Plato's use of fallacy is one which is 
unusually closely connected with dialectic and hence with the 
dialogue form, since ( i ) it is in the process of conversation that 
the various instances of fallacious argument almost invariably 
occur, and (2) fallacy is often part of the eXeyx ?, of the 
dialectical shock-treatment administered by Socrates as the 
torpedo-fish. It is for this reason that I have tried to study 
Plato's fallacies as they appear in complete dialogues or at least 
in fairly complete conversations, and also why I have tried to 
take into account the aims and capacities of the various speakers 
Plato has presented to us. 

The chief generalization which I would wish to make in 
conclusion is, then, the one which has already been suggested 
in the introduction: that Plato, when he employs fallacious 
argument, does so in order to draw our attention to some funda- 
mental philosophical problem. We have now seen what some 
of these problems are: the meaning of terms like 'good', 
'voluntary', 'learn', the difficulties in the Eleatic either-or, the 
contrast between Eleaticism and the theory of Forms, the 
problems of participation, of imitation, of learning and 
memory, of otherness, of the possibility of knowledge. Such a 
list shows that the study of fallacy in Plato cannot be separated 
from a study of the questions most central to his thought; it is 
therefore not a topic to be neglected. 




in Fr. Bocheriski's brief survey, Ancient Formal Logic (Amsterdam, 
195 1 ), Plato receives credit for the following contributions: (1) 
some remarks on syntax in the Sophist (probably not original with 
Plato) (p. 14) ; (2) an elementary semantics in the Cratylus (probably 
also not original) (p. 15); (3) an analogon of r Ax zd ~Ax. zd . "Ax" 1 
at Theaetetus 171 A (p. 16); 42 (4) the first conception and clear 
statement of ' the ideal of valid laws of reasoning ' at Timaeus 47B 
(p. 18) ; (5) a shift in attention 'from the apagogic proofs to positive 
demonstrations of statements attributing a propriety to a subject ' 
which paved the way for the logic of terms (p. 18) ; (6) the method 
of division, which became the origin of the syllogism (p. 18) ; (7) the 
rudiments of almost everything in Aristotle's logic with the exception 
of the analytical syllogism and some related doctrines (p. 18). 
On the other hand, Bocheriski has also this to say of Plato's logic: 

His dialectics appears to us as being a confusion of different sciences and 
different methods. It includes on the one hand the art of disputation, 
metaphysics and logic ; on the other hand Plato does not distinguish between 
formal logic, methodology (of a kind rather akin to that of empirical 
sciences of today) and the intuitive approach to (mostly) axiological 
problems. The reading of his dialogues is almost intolerable to a logician, so many 
elementary blunders are contained in them. It will be enough to mention his 
struggling with the false principle r SaP z> SaP" 1 [e.g. Gorgias 507 A; 
Alcibiades I 126C] 43 or the difficulty he has in grasping that who does 
not admit rSaP 1 must not necessarily admit r SeP n . [E.g. Gorgias 466A 44 ; 

42 See also Formale Logik, Freiburg/Miinchen, 1956, pp. 38-9. 

43 I have looked at Alcibiades I 126C a number of times, but I must 
confess I do not see its relevance in this context. 

44 At Gorgias 466A Polus tries to assert that rhetoric is flattery rather than, 
what Socrates has been at pains to assert, that it is a branch (/xopiov) of 



Meno 73E 45 ; Protagoras 35ofT. 48 ] (p. 17). ... Correct logic we find none in his 
work ; he was, however, a thinker who during his whole life was searching for 
logic and paved the way for its founder (p. 18). (My italics.) 

Bocheriski again discusses Plato's logic in his much larger Formale 
Logik (Freiburg/Munchen, 1956, pp. 39-46). He again credits 
Plato with the first clear concept of logic (using the same Timaeus 
passage as before), 47 and with the method of division (which he 
here illustrates with parts of Sophist 218D-221C). There are no 
remarks about logical blunders of the sort italicized above, but so 
much emphasis is laid upon the immense difficulty which Plato 
experienced in arriving at the most simple logical laws that it is 
hard to feel that Bocheriski wishes to give Plato very much credit in 
cases of his ultimate success : 

Wie schwer ihn die Losung der uns als elementar anmutenden Fragen der 
Logik ankam, mogen die folgenden Stiicke seiner Dialektik nahebringen : er 
ringt in ihnen mit Miihe ura ganz einfache Gesetze (p. 40) . 48 

The illustrative passages are Gorgias 507A, Euthyphro 12A, and 
Protagoras 35oCff. (The Euthyphro passage is new, whereas Gorgias 
466A, Meno 73E, and Alcibiades I 126G have been dropped.) 
Bochenski then writes : 

Im ersten dieser Texte handelt es sich um die (falsche) These : Falls, wenn A 
dem x zukommt, auch B dem x zukommt, dann: falls A dem x nicht 

flattery. There seems to be no question here of anyone wishing to infer r SeP" 1 
from the denial of r SaP n . Rather Plato is simply pointing out that flattery 
is a wider concept than rhetoric. 

45 Meno 73E appears to be similar to Gorgias 466A. Here, when Meno 
says that justice is virtue, Socrates asks, ' Virtue or a virtue ? ' They then agree 
that it must be a virtue since there are other virtues beside justice ; virtue is 
wider than justice, in other words. (Plato does not say so but I would guess 
he means also to imply that the A proposition ' All justice is virtue ' is there- 
fore not convertible simpliciter, and so for 'AH rhetoric is flattery' in the 
Gorgias. See below on Euthyphro i2Aff.) 

46 At Protagoras 350 ff. the question is whether r SaP n implies r PaS n . 
(Plato is clear that it does not.) But see below pp. 94-96. 

47 I should certainly like to see the clear concept of logic which Bochenski 
sees in this passage, but I honestly do not. In commenting on it he writes 
'Ohne den Begriff des allgemeingultigen Gesetzes war offenbar keine 
formale Logik moglich' (p. 40), and with this I should certainly agree. 
But ^oyiafioi 47 C are surely calculations or reasonings, not laws. 

48 'The following extracts from his dialectic, in which he makes a 
laboured approach to quite simple laws, show how difficult he found it to 
solve logical questions that seem elementary to us.' (Translation by Ivo 
Thomas of Formale Logik with the title A History of Formal Logic, Indiana, 
1 96 1, p. 34-) 



zukommt, dann kommt B dem x nicht zu. Der zweite zeigt die Schwierig- 
keiten um die Konvertibilitat des Allsatzes : ob namlich aus ,alle A sind B' 
auch ,alle B sind A 1 folgt. Wie schwer diese Frage Platon fiel, zeigt in 
noch klarerer Weise der dritte Text; er ist zudem deshalb hochst interessant, 
weil Platon darin, um die Ungultigkeit der genannten Konversionsregel 
aufzuweisen, zu komplizierten ausser-logischen Erorterungen (wie etwa zu 
derjenigen iiber die korperliche Starke) seine Zuflucht nimmt (p. 42). 49 

Since the three texts referred to here seem to be the ones by which 
Bochenski sets special store, let us examine each of them in turn. 

(1) Gorgias 507A 

EQ. Aeyco St) on, el r) oa)<f>pcov [sc. 0t>X>7] a-yadrj iariv, r) rovvavriov rfj 
aaxf>povi 7T€7Tovdvla /ca/07 iariv rjv he avr-q r) a<f>p<x)v re koI aKoXaoros. Ildvv 
ye. Kol nijv 6 ye oaxfrpwv rd TipoarjKOvra vpdrroi av kox nepl deovs kcli nepl 
avdpwTTovs' ov yap av oaj<j>povoi to. fir) Trpoay]KOvra TrpaTTWv. 'AvdyKTj ravr' elvai 

Soc. I say, then, that if the temperate soul is good, one that is in the opposite 
state to this sensible one is bad; and that was the senseless and dissolute 
one. Certainly. And further, the sensible man will do what is fitting as 
regards both gods and men; for he could not be sensible if he did what was 
unfitting. That must needs be so. 

The first sentence of this passage certainly appears to state that 
if the temperate soul is good, the soul in the opposite condition 
(i.e., the intemperate soul) is bad. Therefore, as Bochenski writes, 
this 'involves' the false principle that r Ax^>Bx. 3 . ~ Ax z> ~ Bx 1 . 
Bochenski does not explicitly state that the passage should be cited 
as an example of incorrect logic on Plato's part, but the reader 
certainly receives the impression that the quotation was not given 
with the idea of paying the philosopher a compliment. It would 
seem to me that although Bochenski is right to point out that the 
passage has some relevance to the false principle he has indicated, 
there is need to study it at considerably greater length before it 
could become useful to cite it as evidence either in Plato's favor or 
the reverse. (Just how much praise or blame Bochenski attaches to the 

49 'In the first of these texts is involved the (false) thesis: Suppose, if A 
belongs to x, B also belongs to x, then : if A does not belong to x, then B does 
not belong to x. The second shows the difficulties found concerning the 
convertibility of universal affirmative sentences: viz. whether "all B is A" 
follows from "all A is B". The third text shows still more clearly how hard 
Plato felt these questions to be ; it further has the great interest that, to show 
the invalidity of the fore-going rule of conversion, he betakes himself to 
complicated extra-logical discussions — about bodily strength, for instance' 
(trans. Thomas, p. 35). 



passage is not completely clear.) Further study of the passage might 
be conducted along the following lines : 

(a) Is it possible that rj o(x)^pa)v\}jjvxr]\o,y^ri should be translated 
not, 'the temperate soul is good', but 'the temperate soul is the 
good soul ' ? If Plato means that the temperate soul is a member of 
the class of good things, his inference is invalid. If, however, he 
means that 'x is a temperate soul' and i x is a good soul' are 
equivalent, then his inference is correct. (He would have asserted 
v Bx zd Ax 1 as well as r Ax zd Bx 1 ; with this additional assertion, the 
inference r ~Ax zd ^Bx 1 becomes valid.) 

(b) If Plato had added 'the good soul is temperate' his inference 
would have been valid since v Ax zd Bx . Bx zd Ax . zd . ~ Ax zd ^Bx 1 . 
It is possible that Plato regarded this assertion as obvious and thus 
did not trouble to make it ? (By this I mean to suggest that, in Plato, 
logic may well be affected by literary form.) 

(c) Assuming that the inference is invalid, it is possible that Plato 
is conscious of the fallacy and is employing it deliberately? 50 7 A is 
part of a very long argument, and I shall not attempt to study that 
whole argument here. But in view of the general thesis of this book, 
this seems to me a possibility worth considering. 

(d) Again assuming that the inference is invalid, what are we to 
make of the very next sentence in the same paragraph ? For this 
sentence, as far as I can see, is an instance of the correct principle 
v Ax zd Bx. z> . ~Bx zz> ^Ax 1 (if the sensible man does what is 
fitting, then a man who does what is not fitting is not sensible). 
We can surely not be asked to believe that Plato gets an inference 
wrong in one sentence and right in the next — unless of course 
Bochenski means to cite the passage purely as an example of Plato's 
struggles with simple logical laws, or unless we are to assume, which 
is possible, but seems to me unlikely, that Plato may have thought 
both r Ax zd Bx. zz> . ~Ax zd ~Bx 1 and r Ax zd Bx. zd . ~Bx zd 
"Ax 1 . 

(e) Assuming once more that the inference is invalid, what are we to 
make of passages in other dialogues in which Plato makes similar 
inferences correctly ? For instance : 

(i) Hippias Minor 366B 

2JQ. 'Qs iv i<e(f>a\(XLto apa elprjodai, ol ipcvbets cloiv ol oo(f>ol re KOtl hvvarol tfievbeodai. 

in. Ned. 

EQ. 'ASvvaros apa tjjeuBeoOai avrjp Kal apLadrjs ovk av eir) tpcvS'qs. 

III. "E)(€t OVTOiS. 



Soc. In short, then, the false are those who are wise and powerful in uttering 

Hipp. Yes. 
Soc. A man, then, who has not the power to utter falsehoods and is ignorant 

would not be false. 
Hipp. That is true. 

Like the second sentence of the Gorgias passage, this is an instance 
of the correct principle r 'Ax id Bx. zd . ~Bx id ^Ax 1 . 

(ii) Euthyphro 7A 

EY. "Eon Totvvv to {iev to is Oeols Trpoo<f>i\ks ooiov, to 8c lit] Trpoo^iXks avooiov. 

Eu. Well then, what is dear to the gods is holy, and what is not dear to them 
is unholy. 

This passage appears to contain the same mistake as the first sentence 
of the Gorgias passage. However, if we follow the argument along to 
the point of its refutation at ioEff., we shall find that Socrates has 
been regarding Euthyphro's proposed definition as a biconditional 
not as a case of class inclusion. In this case, the inference at 7A 
would have been correct. 

(2) Euthyphro 12A 

£Q. . . . t'Se yap, el ovk avayKalov ooi boKei Sikcciov elvai rrav to ooiov. 
EY. "Enoiye. 


ooiov, aXKa. to pkv ovtov ooiov, to Se ti kcci aX\o; 
EY. Ovx eVo/xai, al Zd)KpaTes, to is Xeyopidvois. 

Soc. . . .Just see whether you do not think that everything that is holy is 

Eu. I do. 
Soc. But is everything that is right also holy? Or is all which is holy right, 

and not all which is right holy, but part of it holy and part something 

Eu. I can't follow you, Socrates. 

Quite far from seeing in this passage evidence that Plato had 
difficulty with the correct conversion of universal propositions, I see 
here evidence that he had no such difficulty but understood per« 
fectly that an A proposition converts not simpliciter, but per ac- 
cidens. Plato's complete mastery of this fact becomes even more 
evident when we see him immediately going on to discuss another A 
proposition, 'all fear is reverence*. This Socrates rejects since he can 
think of cases in which men fear things which they do not reverence. 



(An O proposition contradicts an A proposition, in other words.) 
However, he thinks that the converse, 'all reverence is fear 5 , will 
hold, and this statement is perfectly compatible with 'some fear is 
not reverence'. The reason why these two statements may be 
asserted together whereas it would not be correct to make a joint 
assertion consisting of ' all fear is reverence ' and ' some fear is not 
reverence' could hardly be explained more clearly than it is by 
Plato at 12C: 

ZQ. Ovk ap* opdws e^ei ^*Y €tv ' " va Y&P Beos, €vda Kal alBws' aAA' tva p.ev albws, evda 
Kal Seos, ov (xevTOL tva ye Seos, TravTayov al8u>s. cm TiXeov yap, otfiai, Seos alSovs' 
fiopiov yap alScbs Scot;?, oiorcep api9p,ov irepirrov, wore oi>x tva rrep api6fj.6s, evda Kal 
irepiTTOv, tva Se nepiTrov, evda Kal apidpios. etrei yap ttov vvv ye; 

EY. Ildvv ye. 

Soc. Then it is not correct to say 'where fear is, there also is reverence'. On 
the contrary, where reverence is, there also is fear; but reverence is not 
everywhere where fear is, as I think, fear is more comprehensive than 
reverence ; for reverence is a part of fear, just as the odd is a part of number, 
so that it is not true that where number is there also is the odd, but that 
where the odd is there also is number. Perhaps you follow me now ? 

Eu. Perfectly. 

In other words, Socrates is here saying that where one concept is 
wider (nXeov) than another, it is correct to predicate this wider 
concept of the narrower one but incorrect to predicate the narrower 
concept of the wider. ' Fear ', in this instance, is a wider concept than 
reverence, since although everything which is reverenced is also 
feared, there are some things which are feared but not reverenced. 
When Euthyphro has understood this point, it is then possible for 
Socrates to return to his original point about the holy and the right : 

UQ. To tolovtov tolwv Kal €K€L Xeywv rjpwTcov, apa tva hiKaiov, cvOa Kal oaiov, rj tva 
p.kv oaiov, Zvda Kal hiKaiov, tva Be BiKaiov, ov iravraxov oaiov [xopiov yap rov BiKaiov to 
ooiov. ovtoj <f>d)fj.ev rj aXXo)S 001 80/cei; 

EY. Ovk, aAA' ovtco. <f>alvei yap p.01 6pda>s Xeyeiv. 

Soc. It was something of this sort that I meant before, when I asked whether 
where the right is, there also is holiness, or where holiness is, there also is 
the right; for holiness is not everywhere where the right is, for holiness is 
a part of the right. Do you agree to this, or do you dissent? 

Euth. No, I agree; for I think the statement is correct. (i2Cff.) 

Here it is plain that Plato means to establish the fact that ' right ' is a 
wider concept than 'holy' in the same way that 'fear' is a wider 
concept than 'reverence'. Since he has already pointed out that the 
latter pair is not convertible, and has done this for the sake of 
illustration, it hardly seems possible that he could have thought the 
former pair to be convertible. 

H 93 


However, I am perhaps misunderstanding the inference which 
Bochenski wishes to have drawn from his use oiEuthyphro 12A as an 
illustration of Plato's difficulties with the convertibility of universals; 
perhaps the reader is intended to infer that although Plato found the 
question difficult, his struggles met with success. Even so, the 
general impression given is that if Plato had to work so hard to 
formulate such a very elementary rule, he must have been rather 
particularly stupid. Thus the formulation of the rule, even if 
achieved after this extraordinary expenditure of effort, would seem 
to be regarded as of no great value. 

(3) Protagoras 350G-351B 

IJPQ. . . . eycoye ipcoTrjdeis vtto aov, el ol dvhpeloi OappaXeoi eloiv, cbpioXoyrjoa' el 
he /cat ol OappaXeoi avhpeioi, ovk 'qpcor^drjV el yap p,e Tore rjpov, elirov av on ov 
Trdvres' tovs he dvhpeiovs cos ov OappaXeoi elm, to epiov 6pLoX6yt]jxa ovhap,ov eWSet£as 
cos ovk opOcos cbp,oX6yqoa. eneira tovs emarapievovs ccvtovs eavrcov OappaXecoTepovs 
ovras a.TTO<f>alveis /cat /lit) eVtoTa/ze'vajv aXXcov, /cat ev tovtco oiei rqv avhpeiav /cat tt)v 
oocf>iav ravrov etvar tovtco he tlo Tpotrco pericbv /cat T17V loxyv olr)Qeir)s av etvai oo<f>iav. 
TTpcorov fiev yap el ovtco p.erioov epoio [ie el ol ta^upot Suva-rot elm, cf>air]v av: enena, 
el ol emoTcifjLevoL iraXaieiv hvvaTcoTepoi elm tcov firj €7TiOTap.evcov rraXaleiv /cat au-roi 
avrcov, eVetSav fidOcomv, 77 irplv piaOeiv, <f>air)v av ravra he ep,ov 6p,oXoyrjoavTos 
e^elrj av ooi, xpa>p.€va> to is avTois reap*] plots tovtois, Xeyeiv cos Kara, ttjv ifxrjv 
6p.oXoyiav 17 oo<f>ta eorlv laxvs. eya> be ovhapiov ouS' evravda opioXoyco tovs Bwarovs 
loxvpovs etvai, tovs fxevTOi loxvpovs hvvaTOVs: ov yap raurov etvai hvvapiiv -re /cat 
layvv, dXXa to fxev /cat airo emoTripnjs yiyveoOai, ttjv hvvapnv, /cat coto fiavias ye /cat 
Ovp.ov, loxyv he aVo cf>voecos /cat evTpotfrlas tcov acop.drcov. ovtoo he /ca/cct ov toutov 
etvai ddpoos tc /cat dvhpelav cootc ovp.f}aivei tovs /xev dvhpeiovs OappaXeovs etvai, /ir/ 
pievroi tovs ye OappaXeovs dvhpeiovs irdvTas' ddpoos p-ev yap /cat 077-0 Te^vrj? yiyverai 
dvdpooTTois /cat dvo Ovp,ov ye /cat a7ro piavias, coonep rj hvvapus, dvhpeia he diro <f>voecos 
/cat evTpo<f)ias tcov ifivx<*>v ytyverat. 

Pro. . . . When you asked me whether courageous men are bold, I admitted 
it: I was not asked whether bold men are courageous. Had you asked me 
this before, I should have said — ' Not all. ' And as to proving that coura- 
geous men are not bold, you have nowhere pointed out that I was wrong 
in my admission that they are. Next you show that such persons indivi- 
dually are bolder when they have knowledge, and bolder than others who 
lack it, and therewith you take courage and wisdom to be the same: 
proceeding in this manner you might even take strength to be wisdom. 
On this method you might begin by asking me whether the strong 
are powerful, and I should say ' Yes ' ; and then, whether those who know 
how to wrestle are more powerful than those who do not know how to 
wrestle, and whether individually they are more powerful when they have 
learnt than before learning, and I should say 'Yes'. And on my admitting 
these points it would be open to you to say, by the same token, that 
according to my admission, wisdom is strength. But neither there nor else- 
where do I admit that the powerful are strong, only that the strong are 



powerful ; for I hold that power and strength are not the same, but that 
one of them, power, comes from knowledge, or from madness or rage, 
whereas strength comes from constitution and fit nurture of the body. 
So, in the other instance, boldness and courage are not the same, and there- 
for it results that the courageous are bold, but not that the bold are 
courageous ; for boldness comes to a man from art, or from rage or mad- 
ness, like power, whereas courage comes from constitution and fit nurture 
of the soul. 

This third text shows in even clearer fashion than the second, so 
Bochenski writes, how difficult Plato found the question of whether 
'All A is B' implies 'All B is A'. Just how difficult he found it is 
illustrated by the fact that 'to show the invalidity of the fore- 
going rule of conversion, he betakes himself to complicated extra- 
logical discussions — about bodily strength, for instance' (p. 42). 

Here it seems somewhat more clear that Bochenski thinks that 
Plato has really perceived that A propositions are not convertible 
simpliciter, but because he found the whole question still so difficult 
and because, as a result of his difficulties, he brought in extra-logical 
points, we are apparently again intended to infer that Plato deserves 
no credit for this perception. 

Now, since Plato's remarks on the subject of bodily strength seem 
to be regarded as a particular mark of his difficulties with the 
conversion rule, let us see whether these remarks really are 

The general circumstances in which the passage 35oCff. occurs 
are as follows : in the earlier discussion with Protagoras (329B-334C) 
Socrates attempted (by, as I think, consciously fallacious means [see 
above, p. 28]) to reduce three of the virtues to knowledge, i.e., justice, 
holiness, and temperance. In order to complete the refutation of 
Protagoras (who has maintained 330A that the virtues are distinct) 
it is necessary that the remaining virtue, courage, should be reduced 
to knowledge also. The reduction of courage is more difficult since 
as Protagoras says 349D, 'you will find many people extremely 
unjust, unholy, dissolute, and ignorant, and yet pre-eminently 
courageous'. It is thus postponed until after the discussion of 
Simonides' poem (339A-347A). When the argument is resumed at 
349D, Socrates first attempts a very rapid completion of it at 
349E-350G. After Protagoras rightly points out the false conversion 
which has been used, however, Socrates gives up this line of approach 
(very abruptly, in fact, at 35 iB) and embarks on a longer route to 
his objective, one which involves him in the discussion of pleasure 
which is the distinctive feature of the latter part of the dialogue. 

At the moment we need to examine Socrates' argument 349E- 
350C (the attempted brief reduction of courage to knowledge) and 



to compare it with the remarks on bodily strength in Protagoras' 
speech of correction 350G-351B. We shall then see that what Protag- 
oras has done is to give an argument precisely similar in structure 
to that of Socrates but one which is instantiated by different terms : 


(a) All the courageous are bold. (a) All the strong are powerful. 

(b) those who know (e.g. how to (b) those who know (e.g. how to 
dive into wells) are bolder than wrestle) are more powerful than 
those who do not know, or those who do not know, or 

All the knowing are bold. All the knowing are powerful. 

(c) Some bold men are not (c) (Some powerful men are not 
courageous. strong.) 50 

(d) All the bold are courageous. (d) All the powerful are strong. 

(e) All the knowing are courageous, (e) All the knowing are strong. 
(/) All knowledge is courage. (/) All knowledge is strength. 

The false conversion of (a) occurs at (d) . (d) is then combined with 
(b) to produce, by a valid syllogism, the conclusion (e). (Whether 
(/) is the legitimate result of (e) is a debatable point, but not one 
which we need to consider here.) What Protagoras has done is, in 
fact, very similar to what Socrates has done at Euthyphro i2Aff. 
Euthyphro at first failed to understand what Socrates said about the 
holy and the right, but did understand when Socrates constructed a 
similar argument about reverence and fear. Protagoras shows, in the 
speech 35oCff., that he understands that Socrates' argument contains 
a false conversion, and he makes his understanding clear by con- 
structing a similar argument with different terms. Bocheiiski has 
not quoted enough of the important passage ; if we begin at 349E 
instead of 350C, we can see the parallelism I have just indicated. 
It then becomes clear that Plato's remarks about bodily strength 
are by no means extra-logical, unless of course one wants to say that 
all instantiation of logical laws is extra-logical. (In this case Bocheri- 
ski's objection would apply to the Gorgias and Euthyphro passages as 

All in all, it is very difficult to avoid the conclusion that Fr. 
Bochenski's real quarrel with Plato is that this unenlightened 

60 This step does not occur in the text but has been added for the sake of 
filling out the scheme. I have also reworded some of the other steps (notably 
(b)) in order to show more clearly what I take to have been Plato's line of 



philosopher did not employ the notation of Principia Mathematica or 
something comparable. (In this respect he resembles Robinson, who 
thinks that Plato cannot have understood the importance of equi- 
vocation since he did not call it by a technical name [see above 
p. 7n.].) Plato's understanding of the conversion rule is too obvious 
for Bocheriski to deny it outright, but he cannot, apparently, bring 
himself to congratulate Plato on the correct formulation of this 
rule since it is nowhere stated in the dialogues in technical form. 
Thus he confuses the issue with his talk about 'struggles' and 
'difficulties', in this way contriving to give the impression that 
although some of these struggles may have met with success, Plato 
still remains a second- or third-rate logician. (The possibility that 
some of Plato's ' struggles ' and ' difficulties ' may have a connection 
with his dramatic technique has not been considered.) 

In general I find myself in substantial agreement with what Mr. 
John Ackrill has written in his review of Ancient Formal Logic in Mind, 
LXII, 1953, p. in: 

Fr. Bocheriski's dismissal of Plato in two paragraphs is grotesquely cavalier. 
. . . The passages he refers to are quite inadequate to support such a 
sweeping stricture, even waiving the fact that in some of them Plato is 
exposing not committing fallacies. . . . Plato often uses arguments long and 
short which can readily be symbolised. He does not often state or discuss 
logical laws, but the laws in accordance with which he argues can be 

In Formate Logik, as I have indicated, the stricture is not quite so 
sweeping, and the dismissal takes somewhat longer, but Fr. 
Bochenski still seems to be of the same mind in the matter of Plato's 
logical competence. I hope I have succeeded in demonstrating that 
the question is at least open to somewhat more lengthy debate. 51 

51 In my symbolic notation I have followed Bochenski and have regularly 
omitted quantifiers. 


General Index 

accusative and infinitive construc- 
tion, and fallacy, 25, 30 

Ackrill, J., on Bochenski, 97 

Allan, D. J.: on Cratylus, 54 n. ; on 
Republic I, 72 n. 

Allen, R. E., on predication, 27n. 

ambiguity, see equivocation 

argument, immoral practise in, 

Aristotle: on equivocation, 4m; on 
Hippias Minor, 79 n. ; on how to be 
a sophist, 81 n.; his logic and 
Plato's, 88 ; on refutation, 74 ; on 
secundum quid, 6n.; on solution of 
fallacies, 78 n.; on voluntary 
error, 79 n. 

jSaaiAi/o) Texv-q, 21 

becoming, neglect of, associated 
with certain fallacies, 13, 17, 38, 
40, 42; and Koivwvla, 42 n. 

Booth, N. B., on predication, 27 n. 

Bochenski, I. M., on Plato's logic, 
xiv, 88-97 

Bluck, R. S., on predication, 27 n. 

Charmides, paradox concerning tem- 
perance in, 2 1 

Cherniss, H., reply to Allan, 54 n. 

composition, fallacy of, 49 n. 

convention theory, see language 

contradiction : argument against, 
17, 17m, 23; and equivocation, 

conversion of propositions : in Euthy- 
phro, 92-94; in Protagoras, 95-96 

copula, 'is' used as, 12, 17, 48 
correct speech: and eristic, 33; 

importance of, 5, 8, n 
courage, in Protagoras, 28 n., 84, 


'courage', equivocal in Laches, 

Cratylus, the historical, 54 n. 

Cratylus: 46-64; argument against 
false speaking in, xiii, 47, 63; 
and correct speech, 33; and 
Euthydemus, xiii, 30, 47, 51 ; main 
purpose of, 56-57; and Phaedo, 
56 n.; and Republic I, 47 n.; and 
Sophist, 48 n. ; and Theaetetus, 43 n. ; 
and theory of Forms, xiv, 61, 64 ; 
two theories of language in, 42 n.; 
two versions of natural rightness 
theory in, 51 n., 53; structure of, 
6 in.; true and false names in, 

daimonion, 1 

Seivos, equivocal in Republic I, 72 n. 

deLaguna, T. A., on predication, 

dialectic, and eristic, 2, 3, 12, 21, 

32, 33 
dialogue form, and fallacy, xiv, 

86-87, 9 1 
'different': equivocal, 26; as an 

illegitimate Form, 26-29; and 

secundum quid, 26 
division, fallacy of, 49 n. 
Dodds, E. R., on equivocation, 80 n. 
bwara opav, fallacy using, 25, 30 



Eleatic logic: and Platonic meta- 
physics, 14; and Eleatic monism, 

Eleatic monism: attacked by Plato 
in the Republic, 14m ; its clash with 
Plato's position, xiii, 13, 30 n., 64; 
and equivocation, i6n., 25; in- 
compatibility with the theory of 
Forms, 29, 29 n., 30 n., 87 

Eleatic philosophy: and Heracli- 
teanism, 38, 42, 42 n. ; and images, 
48, 48 n.; and Platonic meta- 
physics, 86, 87 

ZXeyXos: and Aristotle, 74; and 
fallacy, 87 

equivocation : xii, 80 ; combined with 
secundum quid, 6, 12, 13, 17; and 
contradiction, 36; and Eleatic 
monism, i6n.; in Hippias Minor, 
65; and possessive adjectives, 24, 
25, 30; Robinson on, 7m, 8n.; 
and Socratic ethics, 77, 78; used 
to attack theory of Forms, 29 

eristic: not an art, 8; and correct 
speech, 33; and dialectic, 2, 3, 11, 
12, 21, 32, 33; and philosophy, 
32; and teaching, 17, 20; and 
theory of Forms, 33 

* eristic ', as applied to arguments not 
used contentiously, 37 n., 81 

etymology: in Cratylus, 57, 58, 60; 
in other dialogues, 61 n. 

e£ irpoLTTeiv: 8; equivocal, 10; 
Shorey on, ion.; synonyms for, 

Euthydemus, and Protagoras, 50-51 

Euthydemus: 1-33; consciousness of 
fallacy in, xii; and C atylus, xii, 
xiii, 47, 51, 65; and Hippias 
Minor, xiii, 65, 78; and Meno, 
85; and Protagoras, 28 n.; and 
Sophist, xiv, 13, 17m, 27 n.; struc- 
ture of, 1 ; and Theaetetus, xii, 35, 
37 m, 41,42 m, 43, 44, 45 

Euthyphro, conversion of propositions 
in, 92-94 

exhortation, 5, 12, 20, 21, 22 

existential 'is', 13, 17, 48 

fallacies: composition, 49 n.; divi- 
sion, 49 n.; plurium interrogationum, 
4m; see also equivocation, secun- 
dum quid 

fallacy: consciousness of, xi-xiii, 33, 
80, 81, 82, 86; constructive use of, 
28m, 43, 44, 45, 65, 78, 84; and 
dialogue form, 86-87 5 an d cXeyxos, 
87; and philosophy, xi-xiii, 13, 
14, 29, 65, 86, 87; study of 
recommended, xv, 87 

false names, Plato's defense of, 48, 

false opinion: as dAAoSofm, 42; and 
false speaking, 18, 41; Plato's 
defense of, 42; in Theaetetus, 40, 
47 n. 

false speaking: argument against, 
14, 15, 16, 18, 47, 48, 53, 56; and 
Eleaticism, 48 n., 63; and false 
opinion, 41; and ignorance, 19; 
and teaching, 18, 19, 22; and 
theory of Forms, 53 

'folly', equivocal in Protagoras, 28 n. 

Gifford, E. H. : on equivocal use of 

Xeyew, 15m; on equivocal use of 

os, 13m 
'good', equivocal in Hippias Minor, 

65, 68, 74, 76, 77, 87 
good fortune, and wisdom, 8, 1 1 
goods: right use of, 9; must be 

beneficial, 9, 1 1 

Heraclitus (or Heracliteans) : and 
fallacy, 38 ; and names, 6 1 ; and 
Parmenides, 47 n., 52, 54 n.; and 
Protagoras, 34, 38, 47 n. 

Hippias Minor : 65-79 ; and ambigui- 
ties in Socratic ethics, xii, 78; 
Aristotle on, 79 n.; authenticity 
of, 79 n.; and Euthydemus, xii, 65, 
78; and Theaetetus, 78 

Homer: in Cratylus, 53; in Hippias 
Minor, 66, 67, 70 ; in Theaetetus, 34 

images: and Eleaticism, 48 n.; and 
false names, 48; in Sophist, 48 n.; 
and true names, 60 



imitation: of Forms, 63; and lan- 
guage, 62, 63; of reality, 59 

'is', equivocal, 12, 13, 15, 15m, 16, 

Isocrates, in Euthydemus, 31 

KaKws Xeyeiv, equivocal, 16 

Kirk, G. S., on Gratylus, 54 n. 

'know', equivocal, 35, 35 n., 38, 39 

knowing and not knowing: in 
Euthydemus, 6, 23, 24; and secun- 
dum quid, 73; a sophistic rule, 39; 
as a source of fallacy, 39, 40, 41 , 44 

knowledge: and the good, 21-22; 
paradoxes concerning, 21, 56m, 
85; as perception and its connec- 
tion with fallacy, 346°. 

Koivcovta, and becoming, 42 n. 

KpeiTTUiv, equivocal in Gorgias and 
Republic I, 84 n. 

Laches, use of fallacy in, 82-84 
language : convention theory of, 50, 
51, 56, 62, 63; Plato's own theory 
of, 62-64; natural Tightness theory 
of, 50, 5i, 5i n., 52, 53, 56, 62; 
study of not an index to reality, 
54m, 55m, 59, 60 
'learn', equivocal, 5, 6, 7m, i6n., 

29, 35 n., 44, 65, 87 

learning: not accounted for by 
definition of knowledge as per- 
ception, 38, 40, 44, 45 ; and wis- 
dom, 8 

logic: Bochenski on Plato's, xiv, 
88-97; P^to and, xiv, 80, 80 n., 

'mean', equivocal, 19, 20 

memory, not accounted for by 
definition of knowledge as per- 
ception, 36, 39, 40, 44, 45, 56 n. 

Meno: and Cratylus, 86; and Euthy- 
demus, 1 2 n., 85 ; paradox of, 84-86, 
86 n. ; and Theaetetus, 85 

Meridier, L., on Cratylus, 61 n. 

mistakes, possibility of, 18, 19 

names: all are correct, 46, 47, 48, 
53, 53 n -, 62, 64; correct by 
convention, 50 ; true and false, 49, 
49 n., 50, 51 ; true, how identified, 

naming and painting : in Cratylus, 46, 
55, 56; in Phaedo, 56 n. 

natural Tightness theory, see lan- 

opposition, and refutation, 17, 66, 

7rai8ia, 5, 7m 

painting, see naming 

paradox : in Euthydemus, 21; in 

Hippias Minor, 66, 76, 78; in 

Meno, 84-85 
Parmenides: and eristic, i8n., 33; 

and fallacy, xiii, 13; and Hera- 

clitus, 47 n., 54m; his philosophy 

incompatible with the theory of 

Forms, 14m, 29, 29m, 33; 

Plato's attitude towards, i8n. 
Parmenides, incompatibility of Elea- 

ticism and theory of Forms in, 

particulars : and Forms, 25-29, 30 n. ; 

and non-being, 13 
Peck, A. L. : on Euthydemus and 

Sophist, xiv, 27 n.; on Parmenides, 

perception, as a definition of know- 
ledge, 346°. 
Phaedo: and Cratylus, 56 n.; and 

theory of Forms, 85 
Phillips, B., on Meno, 86 n. 
philosopher, and politician, 31, 32 
philosophy : and sophists, 29-30, 3 1 ; 

two varieties in Euthydemus, 32; 

and wisdom, 20 
'powerful': equivocal in Hippias 

Major, 67, 69 n.; in Hippias Minor, 


Proclus, and Hippias Minor, 75 n. 

Protagoras: and the convention 
theory, 50; and Euthydemus, 50, 
51,52; and false speaking, 1 8 ; and 



knowledge as perception, 34; and 
man is the measure, 19, 34, 42, 
50; and Theaetetus, 19 
Protagoras: 94-96; consciousness of 
fallacy in, 28 n.; conversion in, 
95-96; courage in, 84; and 
Euthydemus, 28 n.; self-predication 
in, 28n. 

recollection, doctrine of: in Meno, 
85; in Phaedo, 56 n. 

reflexivity, of argument, 19 

refutation: and false speaking, 18, 
19; of Hippias, 66ff.; and opposi- 
tion, 17, 66, 73, 74; real and 
apparent, 5, 7; and teaching, 19 

Robinson, R. : and Bochenski, 97; 
and Euthydemus, 7m, 8n.; on 
Cratylus, 49 n. 

secundum quid: combined with equi- 
vocation, xii, 6, 12, 13, 17, 48, 80; 
converse of, 6n., 25; defined, 6; 
described by Aristotle, 6n.; and 
'different', 26; in Theaetetus, 39 

self-predication, 27, 27 n., 28m, 30 

sentences, true and false, 49 n. 

Shorey, P.; on Republic I, ion., 
84 n.; on Hippias Minor, 76 n. 

shuttle, ideal, 52 

aiywvra Xeyeiv, fallacy using, 25, 30 

Socrates, the historical, 2 n. 

Socratic ethics, and equivocation, 

77, 78 
Sophist: and Cratylus, 48 n.; and 

Euthydemus, xiv, 13, 17m, 27 n.; 
and Theaetetus, 42 n. 
synonyms, use of by sophists, 10 

Taylor, A. E., on Hippias Minor, 
75m, 76m 

Theaetetus: 34-35; constructive use 
of fallacy in, 44, 45 ; and Cratylus, 
43 n.; and Euthydemus, xii, 35, 
37m, 39, 41, 42m, 43, 44, 45; 
and false opinion, 40-42; and 
false speaking, 41 ; and Meno, 85; 
Protagoras in, 34, 37, 38; role of 
Socrates in, 43-45 

theory of Forms : and Cratylus, xiv, 
45 ? 57> 61, 62, 64; and Eleatic 
philosophy, 14, 14m, 29, 29 n., 
30 n.; and error, 52; and Euthy- 
demus, 25ff. ; and Phaedo, 85 

Thompson, E. S., on Meno and 
Euthydemus, 12m 

virtue: and knowledge, 8; teach- 
ability of, 2, 11, 17, 33; and 
wisdom, 2n., 12, 52 

'voluntary', equivocal in Hippias 
Minor, 65, 75, 76, 77, 87 

Vlastos, G. : on Protagoras, 28 n.; on 
self-predication, 28 n. 

wisdom : and good fortune, 8, 1 1 ; 
and happiness,. 2n., 9, 11; and 
ignorance, 9, 11, 13; teachability 
of, 9, 11, 12, I2n., 1 7 ; and virtue, 
2n., 12, 52 


Index of Passages 



de Interp. 
16a i3ff. 

49 n. 








1025a 6 ff. 




N. Ethics 



1136a iff. 

79 n - 





On Soph. Ref. 



1 66b 38 ff. 




1 68a 37 ff. 




172b ioff. 

81 n. 


47, 51 n., 86 

174a i7ff. 

81 n. 



175a 5ff. 



49 n. 

175a 32 ff. 


43 1 A 


175b 30 ff. 



49 n. 

i75b 39 ff- 







Alcibiades I 




88, 88 n., 89 




1 in. 




60, 63 




3 8 3 Aff. 


438 A ff. 

57-58, 60 


49, 49 n - 


48, 59 


49 n. 



3 8 5 Eff. 



58, 59 



439 A ff 


3 86Bff. 


4 39 Cff. 

57, 60-61 


5°> 51 > 52 



Ep. VII 


5<>> 52, 53, 62 






273 d 






















2 8 4 Cff. 




2 86Dff. 







293 E 

294 ff. 





2 9 9Bff. 









301 ff. 








1, 3 



2, 2n., 11 








30 n. 

















307 A ff. 


6, 35, 44 

5, 711., 20 



7 A 


7 D 







89, 8911., 92, 94, 96 




2n., I2n. 

2n., 11, 20 




88, 88 n., 89, 8911. 

12, 13 


80 n. 





4 89Bff. 

84 n. 



88, 89, 90-92, 96 


Hippias Major 


69 n. 

l 9> 47 


69 n. 

l 9 


69 n. 


Hippias Minor 







22, 23 








68, 79 n. 



68, 73, 79 n - 













6n., 25 


66, 67, 69 





373 Bff. 


3, 25 













33 8Cff. 





72, 75 






66, 73 





























89, 89 n. 

5611., 84-85, 86 n. 



241 Dff. 

2 54 Dff. 




262 ff. 












61 n. 










3 2 9 Bff. 




2811., 95 







i6 3 ff. 


28 n. 

1 63 A 
















350 ff. 

89, 89 n. 


35 oCff. 

89, 94-95, 96 






1 in. 



61 n. 



47 n. 

61 n. 

48 n. 
48 n., 42 


42 n. 
42 n. 

49 n. 

42 n. 



38, 43 n. 




36, 39 














1 89 A 









1 99 A 












!, 4711. 

in Remp. 


", 355K 


61 n. 


89 n. 




3 r ««™»B""»i p£L.%